{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.997275204359674,
  "eval_steps": 500,
  "global_step": 6222,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 1.169871794871795e-05,
      "loss": 0.9422,
      "step": 73
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.33974358974359e-05,
      "loss": 0.8189,
      "step": 146
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.5096153846153845e-05,
      "loss": 0.5106,
      "step": 219
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.67948717948718e-05,
      "loss": 0.4002,
      "step": 292
    },
    {
      "epoch": 0.99,
      "learning_rate": 5.849358974358975e-05,
      "loss": 0.3432,
      "step": 365
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.019230769230769e-05,
      "loss": 0.2873,
      "step": 438
    },
    {
      "epoch": 1.4,
      "learning_rate": 8.189102564102564e-05,
      "loss": 0.2576,
      "step": 511
    },
    {
      "epoch": 1.59,
      "learning_rate": 9.35897435897436e-05,
      "loss": 0.2532,
      "step": 584
    },
    {
      "epoch": 1.79,
      "learning_rate": 9.941228851291185e-05,
      "loss": 0.2378,
      "step": 657
    },
    {
      "epoch": 1.99,
      "learning_rate": 9.811219946571684e-05,
      "loss": 0.2038,
      "step": 730
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.681211041852181e-05,
      "loss": 0.1889,
      "step": 803
    },
    {
      "epoch": 2.39,
      "learning_rate": 9.55120213713268e-05,
      "loss": 0.1846,
      "step": 876
    },
    {
      "epoch": 2.59,
      "learning_rate": 9.421193232413179e-05,
      "loss": 0.1679,
      "step": 949
    },
    {
      "epoch": 2.79,
      "learning_rate": 9.291184327693678e-05,
      "loss": 0.1636,
      "step": 1022
    },
    {
      "epoch": 2.99,
      "learning_rate": 9.161175422974176e-05,
      "loss": 0.1592,
      "step": 1095
    },
    {
      "epoch": 3.19,
      "learning_rate": 9.031166518254676e-05,
      "loss": 0.1301,
      "step": 1168
    },
    {
      "epoch": 3.39,
      "learning_rate": 8.901157613535174e-05,
      "loss": 0.1362,
      "step": 1241
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.771148708815673e-05,
      "loss": 0.1425,
      "step": 1314
    },
    {
      "epoch": 3.79,
      "learning_rate": 8.64113980409617e-05,
      "loss": 0.1363,
      "step": 1387
    },
    {
      "epoch": 3.99,
      "learning_rate": 8.511130899376671e-05,
      "loss": 0.1166,
      "step": 1460
    },
    {
      "epoch": 4.19,
      "learning_rate": 8.381121994657169e-05,
      "loss": 0.1059,
      "step": 1533
    },
    {
      "epoch": 4.39,
      "learning_rate": 8.251113089937668e-05,
      "loss": 0.1111,
      "step": 1606
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.121104185218165e-05,
      "loss": 0.104,
      "step": 1679
    },
    {
      "epoch": 4.78,
      "learning_rate": 7.991095280498666e-05,
      "loss": 0.1116,
      "step": 1752
    },
    {
      "epoch": 4.98,
      "learning_rate": 7.861086375779163e-05,
      "loss": 0.1095,
      "step": 1825
    },
    {
      "epoch": 5.19,
      "learning_rate": 7.731077471059662e-05,
      "loss": 0.0888,
      "step": 1898
    },
    {
      "epoch": 5.38,
      "learning_rate": 7.60106856634016e-05,
      "loss": 0.0966,
      "step": 1971
    },
    {
      "epoch": 5.58,
      "learning_rate": 7.47105966162066e-05,
      "loss": 0.0875,
      "step": 2044
    },
    {
      "epoch": 5.78,
      "learning_rate": 7.341050756901158e-05,
      "loss": 0.0828,
      "step": 2117
    },
    {
      "epoch": 5.98,
      "learning_rate": 7.211041852181657e-05,
      "loss": 0.0984,
      "step": 2190
    },
    {
      "epoch": 6.18,
      "learning_rate": 7.081032947462156e-05,
      "loss": 0.0777,
      "step": 2263
    },
    {
      "epoch": 6.38,
      "learning_rate": 6.951024042742655e-05,
      "loss": 0.0786,
      "step": 2336
    },
    {
      "epoch": 6.58,
      "learning_rate": 6.821015138023152e-05,
      "loss": 0.0751,
      "step": 2409
    },
    {
      "epoch": 6.78,
      "learning_rate": 6.691006233303651e-05,
      "loss": 0.0762,
      "step": 2482
    },
    {
      "epoch": 6.98,
      "learning_rate": 6.56099732858415e-05,
      "loss": 0.0809,
      "step": 2555
    },
    {
      "epoch": 7.18,
      "learning_rate": 6.43098842386465e-05,
      "loss": 0.0668,
      "step": 2628
    },
    {
      "epoch": 7.38,
      "learning_rate": 6.300979519145147e-05,
      "loss": 0.0673,
      "step": 2701
    },
    {
      "epoch": 7.58,
      "learning_rate": 6.170970614425646e-05,
      "loss": 0.0653,
      "step": 2774
    },
    {
      "epoch": 7.78,
      "learning_rate": 6.040961709706144e-05,
      "loss": 0.0662,
      "step": 2847
    },
    {
      "epoch": 7.98,
      "learning_rate": 5.910952804986644e-05,
      "loss": 0.0663,
      "step": 2920
    },
    {
      "epoch": 8.18,
      "learning_rate": 5.7809439002671416e-05,
      "loss": 0.0529,
      "step": 2993
    },
    {
      "epoch": 8.38,
      "learning_rate": 5.650934995547641e-05,
      "loss": 0.0529,
      "step": 3066
    },
    {
      "epoch": 8.57,
      "learning_rate": 5.520926090828139e-05,
      "loss": 0.0585,
      "step": 3139
    },
    {
      "epoch": 8.77,
      "learning_rate": 5.3909171861086385e-05,
      "loss": 0.0538,
      "step": 3212
    },
    {
      "epoch": 8.97,
      "learning_rate": 5.260908281389136e-05,
      "loss": 0.0633,
      "step": 3285
    },
    {
      "epoch": 9.17,
      "learning_rate": 5.130899376669636e-05,
      "loss": 0.0477,
      "step": 3358
    },
    {
      "epoch": 9.37,
      "learning_rate": 5.0008904719501335e-05,
      "loss": 0.0469,
      "step": 3431
    },
    {
      "epoch": 9.57,
      "learning_rate": 4.8708815672306325e-05,
      "loss": 0.0482,
      "step": 3504
    },
    {
      "epoch": 9.77,
      "learning_rate": 4.740872662511131e-05,
      "loss": 0.0528,
      "step": 3577
    },
    {
      "epoch": 9.97,
      "learning_rate": 4.61086375779163e-05,
      "loss": 0.0458,
      "step": 3650
    },
    {
      "epoch": 10.17,
      "learning_rate": 4.480854853072128e-05,
      "loss": 0.0401,
      "step": 3723
    },
    {
      "epoch": 10.37,
      "learning_rate": 4.350845948352627e-05,
      "loss": 0.04,
      "step": 3796
    },
    {
      "epoch": 10.57,
      "learning_rate": 4.2208370436331254e-05,
      "loss": 0.0439,
      "step": 3869
    },
    {
      "epoch": 10.77,
      "learning_rate": 4.0908281389136244e-05,
      "loss": 0.0407,
      "step": 3942
    },
    {
      "epoch": 10.97,
      "learning_rate": 3.9608192341941234e-05,
      "loss": 0.0399,
      "step": 4015
    },
    {
      "epoch": 11.17,
      "learning_rate": 3.830810329474622e-05,
      "loss": 0.0355,
      "step": 4088
    },
    {
      "epoch": 11.37,
      "learning_rate": 3.700801424755121e-05,
      "loss": 0.0356,
      "step": 4161
    },
    {
      "epoch": 11.57,
      "learning_rate": 3.570792520035619e-05,
      "loss": 0.0344,
      "step": 4234
    },
    {
      "epoch": 11.77,
      "learning_rate": 3.440783615316118e-05,
      "loss": 0.0332,
      "step": 4307
    },
    {
      "epoch": 11.96,
      "learning_rate": 3.310774710596616e-05,
      "loss": 0.0356,
      "step": 4380
    },
    {
      "epoch": 12.17,
      "learning_rate": 3.180765805877115e-05,
      "loss": 0.0288,
      "step": 4453
    },
    {
      "epoch": 12.37,
      "learning_rate": 3.0507569011576136e-05,
      "loss": 0.0281,
      "step": 4526
    },
    {
      "epoch": 12.56,
      "learning_rate": 2.9207479964381123e-05,
      "loss": 0.0315,
      "step": 4599
    },
    {
      "epoch": 12.76,
      "learning_rate": 2.790739091718611e-05,
      "loss": 0.0298,
      "step": 4672
    },
    {
      "epoch": 12.96,
      "learning_rate": 2.6607301869991096e-05,
      "loss": 0.0307,
      "step": 4745
    },
    {
      "epoch": 13.16,
      "learning_rate": 2.5307212822796082e-05,
      "loss": 0.0228,
      "step": 4818
    },
    {
      "epoch": 13.36,
      "learning_rate": 2.4007123775601072e-05,
      "loss": 0.0246,
      "step": 4891
    },
    {
      "epoch": 13.56,
      "learning_rate": 2.270703472840606e-05,
      "loss": 0.0247,
      "step": 4964
    },
    {
      "epoch": 13.76,
      "learning_rate": 2.1406945681211045e-05,
      "loss": 0.0268,
      "step": 5037
    },
    {
      "epoch": 13.96,
      "learning_rate": 2.010685663401603e-05,
      "loss": 0.0246,
      "step": 5110
    },
    {
      "epoch": 14.16,
      "learning_rate": 1.8806767586821018e-05,
      "loss": 0.0195,
      "step": 5183
    },
    {
      "epoch": 14.36,
      "learning_rate": 1.7506678539626005e-05,
      "loss": 0.0211,
      "step": 5256
    },
    {
      "epoch": 14.56,
      "learning_rate": 1.620658949243099e-05,
      "loss": 0.02,
      "step": 5329
    },
    {
      "epoch": 14.76,
      "learning_rate": 1.4906500445235978e-05,
      "loss": 0.0212,
      "step": 5402
    },
    {
      "epoch": 14.96,
      "learning_rate": 1.3606411398040964e-05,
      "loss": 0.0212,
      "step": 5475
    },
    {
      "epoch": 15.16,
      "learning_rate": 1.2306322350845949e-05,
      "loss": 0.0172,
      "step": 5548
    },
    {
      "epoch": 15.36,
      "learning_rate": 1.1006233303650935e-05,
      "loss": 0.0176,
      "step": 5621
    },
    {
      "epoch": 15.56,
      "learning_rate": 9.706144256455922e-06,
      "loss": 0.0169,
      "step": 5694
    },
    {
      "epoch": 15.75,
      "learning_rate": 8.406055209260909e-06,
      "loss": 0.0186,
      "step": 5767
    },
    {
      "epoch": 15.95,
      "learning_rate": 7.105966162065895e-06,
      "loss": 0.0177,
      "step": 5840
    },
    {
      "epoch": 16.16,
      "learning_rate": 5.805877114870882e-06,
      "loss": 0.0162,
      "step": 5913
    },
    {
      "epoch": 16.35,
      "learning_rate": 4.505788067675869e-06,
      "loss": 0.0148,
      "step": 5986
    },
    {
      "epoch": 16.55,
      "learning_rate": 3.2056990204808546e-06,
      "loss": 0.0152,
      "step": 6059
    },
    {
      "epoch": 16.75,
      "learning_rate": 1.9056099732858417e-06,
      "loss": 0.0156,
      "step": 6132
    },
    {
      "epoch": 16.95,
      "learning_rate": 6.055209260908283e-07,
      "loss": 0.0144,
      "step": 6205
    }
  ],
  "logging_steps": 73,
  "max_steps": 6239,
  "num_train_epochs": 17,
  "save_steps": 500,
  "total_flos": 5.0512131224764416e+17,
  "trial_name": null,
  "trial_params": null
}