{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7111111111111111,
  "eval_steps": 100,
  "global_step": 100,
  "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
    }
  ],
  "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": 4160649017613312.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}