{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4195555555555557,
  "eval_steps": 100,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0071111111111111115,
      "grad_norm": 147.2163848876953,
      "learning_rate": 0.0,
      "loss": 15.3229,
      "step": 1
    },
    {
      "epoch": 0.014222222222222223,
      "grad_norm": 156.96929931640625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 15.8117,
      "step": 2
    },
    {
      "epoch": 0.021333333333333333,
      "grad_norm": 157.3893280029297,
      "learning_rate": 8.000000000000001e-06,
      "loss": 15.283,
      "step": 3
    },
    {
      "epoch": 0.028444444444444446,
      "grad_norm": 158.99237060546875,
      "learning_rate": 1.2e-05,
      "loss": 15.268,
      "step": 4
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 160.12179565429688,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 15.4867,
      "step": 5
    },
    {
      "epoch": 0.042666666666666665,
      "grad_norm": 156.13417053222656,
      "learning_rate": 2e-05,
      "loss": 15.1959,
      "step": 6
    },
    {
      "epoch": 0.049777777777777775,
      "grad_norm": 169.3893585205078,
      "learning_rate": 1.995215311004785e-05,
      "loss": 15.1078,
      "step": 7
    },
    {
      "epoch": 0.05688888888888889,
      "grad_norm": 180.29098510742188,
      "learning_rate": 1.9904306220095696e-05,
      "loss": 14.6331,
      "step": 8
    },
    {
      "epoch": 0.064,
      "grad_norm": 184.67332458496094,
      "learning_rate": 1.9856459330143542e-05,
      "loss": 14.1344,
      "step": 9
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 184.16485595703125,
      "learning_rate": 1.980861244019139e-05,
      "loss": 13.8721,
      "step": 10
    },
    {
      "epoch": 0.07822222222222222,
      "grad_norm": 206.7319793701172,
      "learning_rate": 1.9760765550239236e-05,
      "loss": 13.7052,
      "step": 11
    },
    {
      "epoch": 0.08533333333333333,
      "grad_norm": 205.46336364746094,
      "learning_rate": 1.9712918660287083e-05,
      "loss": 13.162,
      "step": 12
    },
    {
      "epoch": 0.09244444444444444,
      "grad_norm": 213.31185913085938,
      "learning_rate": 1.966507177033493e-05,
      "loss": 12.6663,
      "step": 13
    },
    {
      "epoch": 0.09955555555555555,
      "grad_norm": 211.15602111816406,
      "learning_rate": 1.9617224880382777e-05,
      "loss": 12.4509,
      "step": 14
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 213.89552307128906,
      "learning_rate": 1.9569377990430624e-05,
      "loss": 12.0618,
      "step": 15
    },
    {
      "epoch": 0.11377777777777778,
      "grad_norm": 224.93186950683594,
      "learning_rate": 1.952153110047847e-05,
      "loss": 11.5434,
      "step": 16
    },
    {
      "epoch": 0.12088888888888889,
      "grad_norm": 209.9617462158203,
      "learning_rate": 1.9473684210526318e-05,
      "loss": 11.2266,
      "step": 17
    },
    {
      "epoch": 0.128,
      "grad_norm": 205.28099060058594,
      "learning_rate": 1.9425837320574165e-05,
      "loss": 11.1783,
      "step": 18
    },
    {
      "epoch": 0.1351111111111111,
      "grad_norm": 208.29965209960938,
      "learning_rate": 1.9377990430622012e-05,
      "loss": 10.1627,
      "step": 19
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 214.2541961669922,
      "learning_rate": 1.933014354066986e-05,
      "loss": 9.8809,
      "step": 20
    },
    {
      "epoch": 0.14933333333333335,
      "grad_norm": 204.05067443847656,
      "learning_rate": 1.9282296650717706e-05,
      "loss": 9.5277,
      "step": 21
    },
    {
      "epoch": 0.15644444444444444,
      "grad_norm": 193.07321166992188,
      "learning_rate": 1.9234449760765553e-05,
      "loss": 8.9317,
      "step": 22
    },
    {
      "epoch": 0.16355555555555557,
      "grad_norm": 186.99464416503906,
      "learning_rate": 1.9186602870813396e-05,
      "loss": 8.929,
      "step": 23
    },
    {
      "epoch": 0.17066666666666666,
      "grad_norm": 186.2172088623047,
      "learning_rate": 1.9138755980861247e-05,
      "loss": 9.0329,
      "step": 24
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 173.13241577148438,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 8.4785,
      "step": 25
    },
    {
      "epoch": 0.18488888888888888,
      "grad_norm": 170.62806701660156,
      "learning_rate": 1.904306220095694e-05,
      "loss": 8.1484,
      "step": 26
    },
    {
      "epoch": 0.192,
      "grad_norm": 171.30633544921875,
      "learning_rate": 1.8995215311004787e-05,
      "loss": 7.3798,
      "step": 27
    },
    {
      "epoch": 0.1991111111111111,
      "grad_norm": 150.02285766601562,
      "learning_rate": 1.894736842105263e-05,
      "loss": 6.6586,
      "step": 28
    },
    {
      "epoch": 0.20622222222222222,
      "grad_norm": 143.904541015625,
      "learning_rate": 1.889952153110048e-05,
      "loss": 6.9233,
      "step": 29
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 148.44595336914062,
      "learning_rate": 1.8851674641148328e-05,
      "loss": 7.3481,
      "step": 30
    },
    {
      "epoch": 0.22044444444444444,
      "grad_norm": 157.87283325195312,
      "learning_rate": 1.8803827751196172e-05,
      "loss": 6.8199,
      "step": 31
    },
    {
      "epoch": 0.22755555555555557,
      "grad_norm": 136.0242462158203,
      "learning_rate": 1.8755980861244022e-05,
      "loss": 6.6569,
      "step": 32
    },
    {
      "epoch": 0.23466666666666666,
      "grad_norm": 145.17214965820312,
      "learning_rate": 1.870813397129187e-05,
      "loss": 5.6681,
      "step": 33
    },
    {
      "epoch": 0.24177777777777779,
      "grad_norm": 138.59141540527344,
      "learning_rate": 1.8660287081339713e-05,
      "loss": 6.2618,
      "step": 34
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 126.07086181640625,
      "learning_rate": 1.8612440191387563e-05,
      "loss": 5.3403,
      "step": 35
    },
    {
      "epoch": 0.256,
      "grad_norm": 134.73928833007812,
      "learning_rate": 1.8564593301435407e-05,
      "loss": 5.3615,
      "step": 36
    },
    {
      "epoch": 0.26311111111111113,
      "grad_norm": 131.0662384033203,
      "learning_rate": 1.8516746411483253e-05,
      "loss": 4.6661,
      "step": 37
    },
    {
      "epoch": 0.2702222222222222,
      "grad_norm": 123.21418762207031,
      "learning_rate": 1.8468899521531104e-05,
      "loss": 5.0741,
      "step": 38
    },
    {
      "epoch": 0.2773333333333333,
      "grad_norm": 121.35318756103516,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 4.7446,
      "step": 39
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 102.97701263427734,
      "learning_rate": 1.8373205741626798e-05,
      "loss": 4.1012,
      "step": 40
    },
    {
      "epoch": 0.29155555555555557,
      "grad_norm": 100.17095947265625,
      "learning_rate": 1.832535885167464e-05,
      "loss": 3.9313,
      "step": 41
    },
    {
      "epoch": 0.2986666666666667,
      "grad_norm": 91.87388610839844,
      "learning_rate": 1.8277511961722488e-05,
      "loss": 3.8289,
      "step": 42
    },
    {
      "epoch": 0.30577777777777776,
      "grad_norm": 92.78990936279297,
      "learning_rate": 1.822966507177034e-05,
      "loss": 3.9649,
      "step": 43
    },
    {
      "epoch": 0.3128888888888889,
      "grad_norm": 93.31304168701172,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 3.8026,
      "step": 44
    },
    {
      "epoch": 0.32,
      "grad_norm": 96.7074203491211,
      "learning_rate": 1.813397129186603e-05,
      "loss": 3.8662,
      "step": 45
    },
    {
      "epoch": 0.32711111111111113,
      "grad_norm": 88.96202850341797,
      "learning_rate": 1.8086124401913876e-05,
      "loss": 3.5804,
      "step": 46
    },
    {
      "epoch": 0.3342222222222222,
      "grad_norm": 83.96373748779297,
      "learning_rate": 1.8038277511961723e-05,
      "loss": 3.4166,
      "step": 47
    },
    {
      "epoch": 0.3413333333333333,
      "grad_norm": 86.4635238647461,
      "learning_rate": 1.799043062200957e-05,
      "loss": 3.2399,
      "step": 48
    },
    {
      "epoch": 0.34844444444444445,
      "grad_norm": 85.60812377929688,
      "learning_rate": 1.7942583732057417e-05,
      "loss": 2.9237,
      "step": 49
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 74.39842224121094,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 3.1092,
      "step": 50
    },
    {
      "epoch": 0.3626666666666667,
      "grad_norm": 78.66775512695312,
      "learning_rate": 1.7846889952153114e-05,
      "loss": 2.9677,
      "step": 51
    },
    {
      "epoch": 0.36977777777777776,
      "grad_norm": 73.33041381835938,
      "learning_rate": 1.7799043062200958e-05,
      "loss": 2.9961,
      "step": 52
    },
    {
      "epoch": 0.3768888888888889,
      "grad_norm": 69.15443420410156,
      "learning_rate": 1.7751196172248805e-05,
      "loss": 2.5786,
      "step": 53
    },
    {
      "epoch": 0.384,
      "grad_norm": 69.43597412109375,
      "learning_rate": 1.770334928229665e-05,
      "loss": 2.6235,
      "step": 54
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 67.97980499267578,
      "learning_rate": 1.76555023923445e-05,
      "loss": 2.9034,
      "step": 55
    },
    {
      "epoch": 0.3982222222222222,
      "grad_norm": 65.86293029785156,
      "learning_rate": 1.7607655502392345e-05,
      "loss": 2.5154,
      "step": 56
    },
    {
      "epoch": 0.4053333333333333,
      "grad_norm": 70.66265106201172,
      "learning_rate": 1.7559808612440192e-05,
      "loss": 2.4679,
      "step": 57
    },
    {
      "epoch": 0.41244444444444445,
      "grad_norm": 67.84262084960938,
      "learning_rate": 1.751196172248804e-05,
      "loss": 2.2115,
      "step": 58
    },
    {
      "epoch": 0.41955555555555557,
      "grad_norm": 71.07489013671875,
      "learning_rate": 1.7464114832535886e-05,
      "loss": 2.3696,
      "step": 59
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 66.5838394165039,
      "learning_rate": 1.7416267942583733e-05,
      "loss": 2.2369,
      "step": 60
    },
    {
      "epoch": 0.43377777777777776,
      "grad_norm": 59.810028076171875,
      "learning_rate": 1.736842105263158e-05,
      "loss": 1.9434,
      "step": 61
    },
    {
      "epoch": 0.4408888888888889,
      "grad_norm": 60.31599426269531,
      "learning_rate": 1.7320574162679427e-05,
      "loss": 2.1169,
      "step": 62
    },
    {
      "epoch": 0.448,
      "grad_norm": 63.39937210083008,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 2.1662,
      "step": 63
    },
    {
      "epoch": 0.45511111111111113,
      "grad_norm": 54.780609130859375,
      "learning_rate": 1.722488038277512e-05,
      "loss": 1.8301,
      "step": 64
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 51.63141632080078,
      "learning_rate": 1.7177033492822968e-05,
      "loss": 1.8786,
      "step": 65
    },
    {
      "epoch": 0.4693333333333333,
      "grad_norm": 61.2618293762207,
      "learning_rate": 1.7129186602870815e-05,
      "loss": 1.6842,
      "step": 66
    },
    {
      "epoch": 0.47644444444444445,
      "grad_norm": 54.38351058959961,
      "learning_rate": 1.7081339712918662e-05,
      "loss": 1.665,
      "step": 67
    },
    {
      "epoch": 0.48355555555555557,
      "grad_norm": 55.260860443115234,
      "learning_rate": 1.703349282296651e-05,
      "loss": 1.6765,
      "step": 68
    },
    {
      "epoch": 0.49066666666666664,
      "grad_norm": 62.19572067260742,
      "learning_rate": 1.6985645933014356e-05,
      "loss": 1.6315,
      "step": 69
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 55.4731559753418,
      "learning_rate": 1.6937799043062203e-05,
      "loss": 1.6022,
      "step": 70
    },
    {
      "epoch": 0.5048888888888889,
      "grad_norm": 53.94541931152344,
      "learning_rate": 1.688995215311005e-05,
      "loss": 1.6217,
      "step": 71
    },
    {
      "epoch": 0.512,
      "grad_norm": 48.46171188354492,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 1.457,
      "step": 72
    },
    {
      "epoch": 0.5191111111111111,
      "grad_norm": 55.64387512207031,
      "learning_rate": 1.6794258373205743e-05,
      "loss": 1.3321,
      "step": 73
    },
    {
      "epoch": 0.5262222222222223,
      "grad_norm": 53.93634033203125,
      "learning_rate": 1.674641148325359e-05,
      "loss": 1.2584,
      "step": 74
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 52.08723449707031,
      "learning_rate": 1.6698564593301437e-05,
      "loss": 1.3642,
      "step": 75
    },
    {
      "epoch": 0.5404444444444444,
      "grad_norm": 62.591835021972656,
      "learning_rate": 1.6650717703349284e-05,
      "loss": 1.2531,
      "step": 76
    },
    {
      "epoch": 0.5475555555555556,
      "grad_norm": 53.1253547668457,
      "learning_rate": 1.6602870813397128e-05,
      "loss": 1.1524,
      "step": 77
    },
    {
      "epoch": 0.5546666666666666,
      "grad_norm": 44.78287124633789,
      "learning_rate": 1.6555023923444978e-05,
      "loss": 1.0058,
      "step": 78
    },
    {
      "epoch": 0.5617777777777778,
      "grad_norm": 48.20844268798828,
      "learning_rate": 1.6507177033492825e-05,
      "loss": 0.9914,
      "step": 79
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 47.6641731262207,
      "learning_rate": 1.6459330143540672e-05,
      "loss": 1.0974,
      "step": 80
    },
    {
      "epoch": 0.576,
      "grad_norm": 52.28041458129883,
      "learning_rate": 1.641148325358852e-05,
      "loss": 1.1361,
      "step": 81
    },
    {
      "epoch": 0.5831111111111111,
      "grad_norm": 51.71315002441406,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 1.1431,
      "step": 82
    },
    {
      "epoch": 0.5902222222222222,
      "grad_norm": 44.117557525634766,
      "learning_rate": 1.6315789473684213e-05,
      "loss": 0.9574,
      "step": 83
    },
    {
      "epoch": 0.5973333333333334,
      "grad_norm": 46.53805160522461,
      "learning_rate": 1.626794258373206e-05,
      "loss": 0.9035,
      "step": 84
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 42.801536560058594,
      "learning_rate": 1.6220095693779903e-05,
      "loss": 0.8866,
      "step": 85
    },
    {
      "epoch": 0.6115555555555555,
      "grad_norm": 42.011192321777344,
      "learning_rate": 1.6172248803827754e-05,
      "loss": 0.6358,
      "step": 86
    },
    {
      "epoch": 0.6186666666666667,
      "grad_norm": 39.56203842163086,
      "learning_rate": 1.61244019138756e-05,
      "loss": 0.6602,
      "step": 87
    },
    {
      "epoch": 0.6257777777777778,
      "grad_norm": 45.57675552368164,
      "learning_rate": 1.6076555023923444e-05,
      "loss": 0.6939,
      "step": 88
    },
    {
      "epoch": 0.6328888888888888,
      "grad_norm": 44.18901443481445,
      "learning_rate": 1.6028708133971294e-05,
      "loss": 0.6864,
      "step": 89
    },
    {
      "epoch": 0.64,
      "grad_norm": 40.97324752807617,
      "learning_rate": 1.5980861244019138e-05,
      "loss": 0.6563,
      "step": 90
    },
    {
      "epoch": 0.6471111111111111,
      "grad_norm": 39.4112663269043,
      "learning_rate": 1.593301435406699e-05,
      "loss": 0.5773,
      "step": 91
    },
    {
      "epoch": 0.6542222222222223,
      "grad_norm": 31.436506271362305,
      "learning_rate": 1.5885167464114835e-05,
      "loss": 0.5103,
      "step": 92
    },
    {
      "epoch": 0.6613333333333333,
      "grad_norm": 44.75907897949219,
      "learning_rate": 1.583732057416268e-05,
      "loss": 0.8296,
      "step": 93
    },
    {
      "epoch": 0.6684444444444444,
      "grad_norm": 41.57793426513672,
      "learning_rate": 1.578947368421053e-05,
      "loss": 0.5342,
      "step": 94
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 50.84144592285156,
      "learning_rate": 1.5741626794258376e-05,
      "loss": 0.5866,
      "step": 95
    },
    {
      "epoch": 0.6826666666666666,
      "grad_norm": 35.748844146728516,
      "learning_rate": 1.569377990430622e-05,
      "loss": 0.43,
      "step": 96
    },
    {
      "epoch": 0.6897777777777778,
      "grad_norm": 34.87845993041992,
      "learning_rate": 1.564593301435407e-05,
      "loss": 0.5604,
      "step": 97
    },
    {
      "epoch": 0.6968888888888889,
      "grad_norm": 30.870040893554688,
      "learning_rate": 1.5598086124401914e-05,
      "loss": 0.4239,
      "step": 98
    },
    {
      "epoch": 0.704,
      "grad_norm": 32.96578598022461,
      "learning_rate": 1.555023923444976e-05,
      "loss": 0.4253,
      "step": 99
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 29.156658172607422,
      "learning_rate": 1.550239234449761e-05,
      "loss": 0.3539,
      "step": 100
    },
    {
      "epoch": 0.7111111111111111,
      "eval_loss": 0.3586403727531433,
      "eval_runtime": 10.0992,
      "eval_samples_per_second": 24.755,
      "eval_steps_per_second": 12.377,
      "step": 100
    },
    {
      "epoch": 0.7182222222222222,
      "grad_norm": 24.950382232666016,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.2712,
      "step": 101
    },
    {
      "epoch": 0.7253333333333334,
      "grad_norm": 30.85600471496582,
      "learning_rate": 1.5406698564593305e-05,
      "loss": 0.274,
      "step": 102
    },
    {
      "epoch": 0.7324444444444445,
      "grad_norm": 26.22223663330078,
      "learning_rate": 1.535885167464115e-05,
      "loss": 0.3479,
      "step": 103
    },
    {
      "epoch": 0.7395555555555555,
      "grad_norm": 36.12468719482422,
      "learning_rate": 1.5311004784688995e-05,
      "loss": 0.3908,
      "step": 104
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 23.735008239746094,
      "learning_rate": 1.5263157894736846e-05,
      "loss": 0.2197,
      "step": 105
    },
    {
      "epoch": 0.7537777777777778,
      "grad_norm": 27.028888702392578,
      "learning_rate": 1.5215311004784689e-05,
      "loss": 0.3645,
      "step": 106
    },
    {
      "epoch": 0.7608888888888888,
      "grad_norm": 25.758739471435547,
      "learning_rate": 1.5167464114832538e-05,
      "loss": 0.1984,
      "step": 107
    },
    {
      "epoch": 0.768,
      "grad_norm": 26.447772979736328,
      "learning_rate": 1.5119617224880383e-05,
      "loss": 0.2023,
      "step": 108
    },
    {
      "epoch": 0.7751111111111111,
      "grad_norm": 24.961519241333008,
      "learning_rate": 1.507177033492823e-05,
      "loss": 0.236,
      "step": 109
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 25.81987953186035,
      "learning_rate": 1.5023923444976079e-05,
      "loss": 0.2197,
      "step": 110
    },
    {
      "epoch": 0.7893333333333333,
      "grad_norm": 18.16558265686035,
      "learning_rate": 1.4976076555023924e-05,
      "loss": 0.1589,
      "step": 111
    },
    {
      "epoch": 0.7964444444444444,
      "grad_norm": 26.138601303100586,
      "learning_rate": 1.4928229665071772e-05,
      "loss": 0.3176,
      "step": 112
    },
    {
      "epoch": 0.8035555555555556,
      "grad_norm": 19.42313003540039,
      "learning_rate": 1.488038277511962e-05,
      "loss": 0.1728,
      "step": 113
    },
    {
      "epoch": 0.8106666666666666,
      "grad_norm": 25.487140655517578,
      "learning_rate": 1.4832535885167465e-05,
      "loss": 0.2062,
      "step": 114
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 22.577999114990234,
      "learning_rate": 1.4784688995215313e-05,
      "loss": 0.2076,
      "step": 115
    },
    {
      "epoch": 0.8248888888888889,
      "grad_norm": 22.75994300842285,
      "learning_rate": 1.4736842105263159e-05,
      "loss": 0.2134,
      "step": 116
    },
    {
      "epoch": 0.832,
      "grad_norm": 20.01392364501953,
      "learning_rate": 1.4688995215311006e-05,
      "loss": 0.1605,
      "step": 117
    },
    {
      "epoch": 0.8391111111111111,
      "grad_norm": 36.355953216552734,
      "learning_rate": 1.4641148325358854e-05,
      "loss": 0.2062,
      "step": 118
    },
    {
      "epoch": 0.8462222222222222,
      "grad_norm": 20.931217193603516,
      "learning_rate": 1.45933014354067e-05,
      "loss": 0.1005,
      "step": 119
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 11.33279800415039,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.0909,
      "step": 120
    },
    {
      "epoch": 0.8604444444444445,
      "grad_norm": 14.551777839660645,
      "learning_rate": 1.4497607655502393e-05,
      "loss": 0.1284,
      "step": 121
    },
    {
      "epoch": 0.8675555555555555,
      "grad_norm": 30.744577407836914,
      "learning_rate": 1.444976076555024e-05,
      "loss": 0.267,
      "step": 122
    },
    {
      "epoch": 0.8746666666666667,
      "grad_norm": 31.37712287902832,
      "learning_rate": 1.4401913875598089e-05,
      "loss": 0.2485,
      "step": 123
    },
    {
      "epoch": 0.8817777777777778,
      "grad_norm": 17.40280532836914,
      "learning_rate": 1.4354066985645934e-05,
      "loss": 0.1856,
      "step": 124
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 19.70085906982422,
      "learning_rate": 1.4306220095693781e-05,
      "loss": 0.1868,
      "step": 125
    },
    {
      "epoch": 0.896,
      "grad_norm": 15.800881385803223,
      "learning_rate": 1.4258373205741626e-05,
      "loss": 0.0932,
      "step": 126
    },
    {
      "epoch": 0.9031111111111111,
      "grad_norm": 20.723295211791992,
      "learning_rate": 1.4210526315789475e-05,
      "loss": 0.1209,
      "step": 127
    },
    {
      "epoch": 0.9102222222222223,
      "grad_norm": 19.325380325317383,
      "learning_rate": 1.4162679425837322e-05,
      "loss": 0.1073,
      "step": 128
    },
    {
      "epoch": 0.9173333333333333,
      "grad_norm": 26.08039665222168,
      "learning_rate": 1.4114832535885167e-05,
      "loss": 0.2808,
      "step": 129
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 42.699058532714844,
      "learning_rate": 1.4066985645933016e-05,
      "loss": 0.1826,
      "step": 130
    },
    {
      "epoch": 0.9315555555555556,
      "grad_norm": 13.087589263916016,
      "learning_rate": 1.4019138755980863e-05,
      "loss": 0.0753,
      "step": 131
    },
    {
      "epoch": 0.9386666666666666,
      "grad_norm": 13.704172134399414,
      "learning_rate": 1.397129186602871e-05,
      "loss": 0.0735,
      "step": 132
    },
    {
      "epoch": 0.9457777777777778,
      "grad_norm": 14.962821006774902,
      "learning_rate": 1.3923444976076557e-05,
      "loss": 0.0665,
      "step": 133
    },
    {
      "epoch": 0.9528888888888889,
      "grad_norm": 8.272432327270508,
      "learning_rate": 1.3875598086124402e-05,
      "loss": 0.0404,
      "step": 134
    },
    {
      "epoch": 0.96,
      "grad_norm": 22.12735366821289,
      "learning_rate": 1.382775119617225e-05,
      "loss": 0.1425,
      "step": 135
    },
    {
      "epoch": 0.9671111111111111,
      "grad_norm": 11.507617950439453,
      "learning_rate": 1.3779904306220097e-05,
      "loss": 0.0626,
      "step": 136
    },
    {
      "epoch": 0.9742222222222222,
      "grad_norm": 17.49174690246582,
      "learning_rate": 1.3732057416267943e-05,
      "loss": 0.0667,
      "step": 137
    },
    {
      "epoch": 0.9813333333333333,
      "grad_norm": 27.13593292236328,
      "learning_rate": 1.3684210526315791e-05,
      "loss": 0.1051,
      "step": 138
    },
    {
      "epoch": 0.9884444444444445,
      "grad_norm": 16.275360107421875,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.1486,
      "step": 139
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 30.099136352539062,
      "learning_rate": 1.3588516746411483e-05,
      "loss": 0.1648,
      "step": 140
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.202574729919434,
      "learning_rate": 1.3540669856459332e-05,
      "loss": 0.0476,
      "step": 141
    },
    {
      "epoch": 1.007111111111111,
      "grad_norm": 12.128777503967285,
      "learning_rate": 1.3492822966507177e-05,
      "loss": 0.0558,
      "step": 142
    },
    {
      "epoch": 1.0142222222222221,
      "grad_norm": 11.825016021728516,
      "learning_rate": 1.3444976076555026e-05,
      "loss": 0.0409,
      "step": 143
    },
    {
      "epoch": 1.0213333333333334,
      "grad_norm": 18.142595291137695,
      "learning_rate": 1.3397129186602873e-05,
      "loss": 0.1332,
      "step": 144
    },
    {
      "epoch": 1.0284444444444445,
      "grad_norm": 16.197526931762695,
      "learning_rate": 1.3349282296650718e-05,
      "loss": 0.0876,
      "step": 145
    },
    {
      "epoch": 1.0355555555555556,
      "grad_norm": 14.817815780639648,
      "learning_rate": 1.3301435406698567e-05,
      "loss": 0.1071,
      "step": 146
    },
    {
      "epoch": 1.0426666666666666,
      "grad_norm": 17.152193069458008,
      "learning_rate": 1.3253588516746412e-05,
      "loss": 0.1171,
      "step": 147
    },
    {
      "epoch": 1.0497777777777777,
      "grad_norm": 24.962247848510742,
      "learning_rate": 1.3205741626794259e-05,
      "loss": 0.1149,
      "step": 148
    },
    {
      "epoch": 1.056888888888889,
      "grad_norm": 9.525614738464355,
      "learning_rate": 1.3157894736842108e-05,
      "loss": 0.0588,
      "step": 149
    },
    {
      "epoch": 1.064,
      "grad_norm": 22.408872604370117,
      "learning_rate": 1.3110047846889953e-05,
      "loss": 0.1329,
      "step": 150
    },
    {
      "epoch": 1.0711111111111111,
      "grad_norm": 10.67056941986084,
      "learning_rate": 1.30622009569378e-05,
      "loss": 0.042,
      "step": 151
    },
    {
      "epoch": 1.0782222222222222,
      "grad_norm": 18.8863525390625,
      "learning_rate": 1.3014354066985647e-05,
      "loss": 0.0745,
      "step": 152
    },
    {
      "epoch": 1.0853333333333333,
      "grad_norm": 12.540481567382812,
      "learning_rate": 1.2966507177033494e-05,
      "loss": 0.0403,
      "step": 153
    },
    {
      "epoch": 1.0924444444444443,
      "grad_norm": 12.860342979431152,
      "learning_rate": 1.2918660287081342e-05,
      "loss": 0.0645,
      "step": 154
    },
    {
      "epoch": 1.0995555555555556,
      "grad_norm": 9.309412956237793,
      "learning_rate": 1.2870813397129188e-05,
      "loss": 0.0406,
      "step": 155
    },
    {
      "epoch": 1.1066666666666667,
      "grad_norm": 12.419631958007812,
      "learning_rate": 1.2822966507177035e-05,
      "loss": 0.0316,
      "step": 156
    },
    {
      "epoch": 1.1137777777777778,
      "grad_norm": 17.204025268554688,
      "learning_rate": 1.277511961722488e-05,
      "loss": 0.0704,
      "step": 157
    },
    {
      "epoch": 1.1208888888888888,
      "grad_norm": 19.311384201049805,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.0957,
      "step": 158
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 4.371336936950684,
      "learning_rate": 1.2679425837320575e-05,
      "loss": 0.0224,
      "step": 159
    },
    {
      "epoch": 1.1351111111111112,
      "grad_norm": 4.176720142364502,
      "learning_rate": 1.263157894736842e-05,
      "loss": 0.0235,
      "step": 160
    },
    {
      "epoch": 1.1422222222222222,
      "grad_norm": 20.444786071777344,
      "learning_rate": 1.258373205741627e-05,
      "loss": 0.108,
      "step": 161
    },
    {
      "epoch": 1.1493333333333333,
      "grad_norm": 24.902677536010742,
      "learning_rate": 1.2535885167464116e-05,
      "loss": 0.0712,
      "step": 162
    },
    {
      "epoch": 1.1564444444444444,
      "grad_norm": 12.11382007598877,
      "learning_rate": 1.2488038277511963e-05,
      "loss": 0.0285,
      "step": 163
    },
    {
      "epoch": 1.1635555555555555,
      "grad_norm": 6.692980766296387,
      "learning_rate": 1.244019138755981e-05,
      "loss": 0.0312,
      "step": 164
    },
    {
      "epoch": 1.1706666666666667,
      "grad_norm": 11.847846984863281,
      "learning_rate": 1.2392344497607655e-05,
      "loss": 0.044,
      "step": 165
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 21.92530632019043,
      "learning_rate": 1.2344497607655504e-05,
      "loss": 0.0886,
      "step": 166
    },
    {
      "epoch": 1.1848888888888889,
      "grad_norm": 21.285676956176758,
      "learning_rate": 1.2296650717703351e-05,
      "loss": 0.144,
      "step": 167
    },
    {
      "epoch": 1.192,
      "grad_norm": 7.421504974365234,
      "learning_rate": 1.2248803827751196e-05,
      "loss": 0.0208,
      "step": 168
    },
    {
      "epoch": 1.199111111111111,
      "grad_norm": 14.405587196350098,
      "learning_rate": 1.2200956937799045e-05,
      "loss": 0.0423,
      "step": 169
    },
    {
      "epoch": 1.2062222222222223,
      "grad_norm": 26.671585083007812,
      "learning_rate": 1.215311004784689e-05,
      "loss": 0.2092,
      "step": 170
    },
    {
      "epoch": 1.2133333333333334,
      "grad_norm": 5.7989091873168945,
      "learning_rate": 1.2105263157894737e-05,
      "loss": 0.0263,
      "step": 171
    },
    {
      "epoch": 1.2204444444444444,
      "grad_norm": 3.7375221252441406,
      "learning_rate": 1.2057416267942586e-05,
      "loss": 0.0164,
      "step": 172
    },
    {
      "epoch": 1.2275555555555555,
      "grad_norm": 11.479283332824707,
      "learning_rate": 1.2009569377990431e-05,
      "loss": 0.0608,
      "step": 173
    },
    {
      "epoch": 1.2346666666666666,
      "grad_norm": 8.047784805297852,
      "learning_rate": 1.196172248803828e-05,
      "loss": 0.0386,
      "step": 174
    },
    {
      "epoch": 1.2417777777777779,
      "grad_norm": 10.923378944396973,
      "learning_rate": 1.1913875598086126e-05,
      "loss": 0.0375,
      "step": 175
    },
    {
      "epoch": 1.248888888888889,
      "grad_norm": 10.642889022827148,
      "learning_rate": 1.1866028708133972e-05,
      "loss": 0.0336,
      "step": 176
    },
    {
      "epoch": 1.256,
      "grad_norm": 9.513157844543457,
      "learning_rate": 1.181818181818182e-05,
      "loss": 0.0379,
      "step": 177
    },
    {
      "epoch": 1.263111111111111,
      "grad_norm": 18.644704818725586,
      "learning_rate": 1.1770334928229666e-05,
      "loss": 0.0491,
      "step": 178
    },
    {
      "epoch": 1.2702222222222221,
      "grad_norm": 13.674433708190918,
      "learning_rate": 1.1722488038277513e-05,
      "loss": 0.0577,
      "step": 179
    },
    {
      "epoch": 1.2773333333333334,
      "grad_norm": 12.674040794372559,
      "learning_rate": 1.1674641148325361e-05,
      "loss": 0.0681,
      "step": 180
    },
    {
      "epoch": 1.2844444444444445,
      "grad_norm": 8.965036392211914,
      "learning_rate": 1.1626794258373206e-05,
      "loss": 0.0271,
      "step": 181
    },
    {
      "epoch": 1.2915555555555556,
      "grad_norm": 8.475702285766602,
      "learning_rate": 1.1578947368421053e-05,
      "loss": 0.0368,
      "step": 182
    },
    {
      "epoch": 1.2986666666666666,
      "grad_norm": 6.760505199432373,
      "learning_rate": 1.15311004784689e-05,
      "loss": 0.0225,
      "step": 183
    },
    {
      "epoch": 1.3057777777777777,
      "grad_norm": 4.455854415893555,
      "learning_rate": 1.1483253588516747e-05,
      "loss": 0.0168,
      "step": 184
    },
    {
      "epoch": 1.3128888888888888,
      "grad_norm": 14.064568519592285,
      "learning_rate": 1.1435406698564594e-05,
      "loss": 0.0671,
      "step": 185
    },
    {
      "epoch": 1.32,
      "grad_norm": 17.76791763305664,
      "learning_rate": 1.1387559808612441e-05,
      "loss": 0.0314,
      "step": 186
    },
    {
      "epoch": 1.3271111111111111,
      "grad_norm": 15.82896900177002,
      "learning_rate": 1.1339712918660288e-05,
      "loss": 0.0553,
      "step": 187
    },
    {
      "epoch": 1.3342222222222222,
      "grad_norm": 9.709717750549316,
      "learning_rate": 1.1291866028708133e-05,
      "loss": 0.0392,
      "step": 188
    },
    {
      "epoch": 1.3413333333333333,
      "grad_norm": 3.827709197998047,
      "learning_rate": 1.1244019138755982e-05,
      "loss": 0.0143,
      "step": 189
    },
    {
      "epoch": 1.3484444444444446,
      "grad_norm": 2.1333112716674805,
      "learning_rate": 1.1196172248803829e-05,
      "loss": 0.0108,
      "step": 190
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 5.342545986175537,
      "learning_rate": 1.1148325358851674e-05,
      "loss": 0.0181,
      "step": 191
    },
    {
      "epoch": 1.3626666666666667,
      "grad_norm": 14.26878547668457,
      "learning_rate": 1.1100478468899523e-05,
      "loss": 0.0416,
      "step": 192
    },
    {
      "epoch": 1.3697777777777778,
      "grad_norm": 12.15977954864502,
      "learning_rate": 1.105263157894737e-05,
      "loss": 0.0491,
      "step": 193
    },
    {
      "epoch": 1.3768888888888888,
      "grad_norm": 14.03559684753418,
      "learning_rate": 1.1004784688995217e-05,
      "loss": 0.0436,
      "step": 194
    },
    {
      "epoch": 1.384,
      "grad_norm": 13.549796104431152,
      "learning_rate": 1.0956937799043064e-05,
      "loss": 0.0348,
      "step": 195
    },
    {
      "epoch": 1.3911111111111112,
      "grad_norm": 14.576817512512207,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.0479,
      "step": 196
    },
    {
      "epoch": 1.3982222222222223,
      "grad_norm": 6.6930623054504395,
      "learning_rate": 1.0861244019138758e-05,
      "loss": 0.0128,
      "step": 197
    },
    {
      "epoch": 1.4053333333333333,
      "grad_norm": 5.9814453125,
      "learning_rate": 1.0813397129186604e-05,
      "loss": 0.0178,
      "step": 198
    },
    {
      "epoch": 1.4124444444444444,
      "grad_norm": 8.173537254333496,
      "learning_rate": 1.076555023923445e-05,
      "loss": 0.0177,
      "step": 199
    },
    {
      "epoch": 1.4195555555555557,
      "grad_norm": 9.597185134887695,
      "learning_rate": 1.0717703349282298e-05,
      "loss": 0.0243,
      "step": 200
    },
    {
      "epoch": 1.4195555555555557,
      "eval_loss": 0.030772214755415916,
      "eval_runtime": 9.3161,
      "eval_samples_per_second": 26.835,
      "eval_steps_per_second": 13.418,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 423,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8305214899064832.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}