{
  "best_metric": 0.948073701842546,
  "best_model_checkpoint": "/tmp/classification_xlnet/checkpoint-6402",
  "epoch": 40.0,
  "eval_steps": 500,
  "global_step": 7760,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_f1": 0.9461194863204914,
      "eval_loss": 0.17429937422275543,
      "eval_runtime": 3.3479,
      "eval_samples_per_second": 1069.914,
      "eval_steps_per_second": 8.363,
      "step": 194
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.9463986599664992,
      "eval_loss": 0.18371470272541046,
      "eval_runtime": 3.3297,
      "eval_samples_per_second": 1075.773,
      "eval_steps_per_second": 8.409,
      "step": 388
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 2.8030216693878174,
      "learning_rate": 2.806701030927835e-05,
      "loss": 0.1681,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.9444444444444444,
      "eval_loss": 0.1891961693763733,
      "eval_runtime": 3.3441,
      "eval_samples_per_second": 1071.144,
      "eval_steps_per_second": 8.373,
      "step": 582
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.9450027917364601,
      "eval_loss": 0.22569966316223145,
      "eval_runtime": 3.3534,
      "eval_samples_per_second": 1068.173,
      "eval_steps_per_second": 8.35,
      "step": 776
    },
    {
      "epoch": 5.0,
      "eval_f1": 0.9422110552763819,
      "eval_loss": 0.26769694685935974,
      "eval_runtime": 3.3297,
      "eval_samples_per_second": 1075.778,
      "eval_steps_per_second": 8.409,
      "step": 970
    },
    {
      "epoch": 5.154639175257732,
      "grad_norm": 0.4933047294616699,
      "learning_rate": 2.61340206185567e-05,
      "loss": 0.0663,
      "step": 1000
    },
    {
      "epoch": 6.0,
      "eval_f1": 0.9369067560022334,
      "eval_loss": 0.291700154542923,
      "eval_runtime": 3.3477,
      "eval_samples_per_second": 1069.976,
      "eval_steps_per_second": 8.364,
      "step": 1164
    },
    {
      "epoch": 7.0,
      "eval_f1": 0.9391401451702959,
      "eval_loss": 0.3442403972148895,
      "eval_runtime": 3.3601,
      "eval_samples_per_second": 1066.048,
      "eval_steps_per_second": 8.333,
      "step": 1358
    },
    {
      "epoch": 7.731958762886598,
      "grad_norm": 0.666480302810669,
      "learning_rate": 2.4201030927835052e-05,
      "loss": 0.0377,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "eval_f1": 0.9416527079843663,
      "eval_loss": 0.33648690581321716,
      "eval_runtime": 3.3304,
      "eval_samples_per_second": 1075.535,
      "eval_steps_per_second": 8.407,
      "step": 1552
    },
    {
      "epoch": 9.0,
      "eval_f1": 0.9410943606923506,
      "eval_loss": 0.33006414771080017,
      "eval_runtime": 3.3696,
      "eval_samples_per_second": 1063.042,
      "eval_steps_per_second": 8.31,
      "step": 1746
    },
    {
      "epoch": 10.0,
      "eval_f1": 0.9427694025683976,
      "eval_loss": 0.34630170464515686,
      "eval_runtime": 3.3697,
      "eval_samples_per_second": 1063.008,
      "eval_steps_per_second": 8.309,
      "step": 1940
    },
    {
      "epoch": 10.309278350515465,
      "grad_norm": 0.21189871430397034,
      "learning_rate": 2.2268041237113402e-05,
      "loss": 0.0259,
      "step": 2000
    },
    {
      "epoch": 11.0,
      "eval_f1": 0.9385817978782803,
      "eval_loss": 0.37461963295936584,
      "eval_runtime": 3.3721,
      "eval_samples_per_second": 1062.24,
      "eval_steps_per_second": 8.303,
      "step": 2134
    },
    {
      "epoch": 12.0,
      "eval_f1": 0.943606923506421,
      "eval_loss": 0.40551692247390747,
      "eval_runtime": 3.3652,
      "eval_samples_per_second": 1064.414,
      "eval_steps_per_second": 8.32,
      "step": 2328
    },
    {
      "epoch": 12.88659793814433,
      "grad_norm": 2.5254008769989014,
      "learning_rate": 2.0335051546391752e-05,
      "loss": 0.0173,
      "step": 2500
    },
    {
      "epoch": 13.0,
      "eval_f1": 0.937465103294249,
      "eval_loss": 0.4070034921169281,
      "eval_runtime": 3.3609,
      "eval_samples_per_second": 1065.787,
      "eval_steps_per_second": 8.331,
      "step": 2522
    },
    {
      "epoch": 14.0,
      "eval_f1": 0.941931881630374,
      "eval_loss": 0.4143279492855072,
      "eval_runtime": 3.3533,
      "eval_samples_per_second": 1068.201,
      "eval_steps_per_second": 8.35,
      "step": 2716
    },
    {
      "epoch": 15.0,
      "eval_f1": 0.9410943606923506,
      "eval_loss": 0.4321683645248413,
      "eval_runtime": 3.3526,
      "eval_samples_per_second": 1068.426,
      "eval_steps_per_second": 8.352,
      "step": 2910
    },
    {
      "epoch": 15.463917525773196,
      "grad_norm": 0.1633172333240509,
      "learning_rate": 1.8402061855670106e-05,
      "loss": 0.014,
      "step": 3000
    },
    {
      "epoch": 16.0,
      "eval_f1": 0.943606923506421,
      "eval_loss": 0.41818127036094666,
      "eval_runtime": 3.3432,
      "eval_samples_per_second": 1071.418,
      "eval_steps_per_second": 8.375,
      "step": 3104
    },
    {
      "epoch": 17.0,
      "eval_f1": 0.9441652707984366,
      "eval_loss": 0.4001573324203491,
      "eval_runtime": 3.3273,
      "eval_samples_per_second": 1076.555,
      "eval_steps_per_second": 8.415,
      "step": 3298
    },
    {
      "epoch": 18.0,
      "eval_f1": 0.9452819653824679,
      "eval_loss": 0.41942211985588074,
      "eval_runtime": 3.3291,
      "eval_samples_per_second": 1075.958,
      "eval_steps_per_second": 8.411,
      "step": 3492
    },
    {
      "epoch": 18.04123711340206,
      "grad_norm": 0.10323987156152725,
      "learning_rate": 1.6469072164948456e-05,
      "loss": 0.0131,
      "step": 3500
    },
    {
      "epoch": 19.0,
      "eval_f1": 0.9438860971524288,
      "eval_loss": 0.42228585481643677,
      "eval_runtime": 3.3201,
      "eval_samples_per_second": 1078.896,
      "eval_steps_per_second": 8.434,
      "step": 3686
    },
    {
      "epoch": 20.0,
      "eval_f1": 0.9441652707984366,
      "eval_loss": 0.43575653433799744,
      "eval_runtime": 3.3587,
      "eval_samples_per_second": 1066.478,
      "eval_steps_per_second": 8.337,
      "step": 3880
    },
    {
      "epoch": 20.61855670103093,
      "grad_norm": 0.34020230174064636,
      "learning_rate": 1.4536082474226805e-05,
      "loss": 0.0107,
      "step": 4000
    },
    {
      "epoch": 21.0,
      "eval_f1": 0.9410943606923506,
      "eval_loss": 0.45651453733444214,
      "eval_runtime": 3.3444,
      "eval_samples_per_second": 1071.038,
      "eval_steps_per_second": 8.372,
      "step": 4074
    },
    {
      "epoch": 22.0,
      "eval_f1": 0.9433277498604131,
      "eval_loss": 0.4561261534690857,
      "eval_runtime": 3.3568,
      "eval_samples_per_second": 1067.075,
      "eval_steps_per_second": 8.341,
      "step": 4268
    },
    {
      "epoch": 23.0,
      "eval_f1": 0.9452819653824679,
      "eval_loss": 0.43663719296455383,
      "eval_runtime": 3.3556,
      "eval_samples_per_second": 1067.473,
      "eval_steps_per_second": 8.344,
      "step": 4462
    },
    {
      "epoch": 23.195876288659793,
      "grad_norm": 0.08574999868869781,
      "learning_rate": 1.2603092783505155e-05,
      "loss": 0.0094,
      "step": 4500
    },
    {
      "epoch": 24.0,
      "eval_f1": 0.9424902289223898,
      "eval_loss": 0.4435234069824219,
      "eval_runtime": 3.3647,
      "eval_samples_per_second": 1064.573,
      "eval_steps_per_second": 8.322,
      "step": 4656
    },
    {
      "epoch": 25.0,
      "eval_f1": 0.9461194863204914,
      "eval_loss": 0.45271530747413635,
      "eval_runtime": 3.3304,
      "eval_samples_per_second": 1075.534,
      "eval_steps_per_second": 8.407,
      "step": 4850
    },
    {
      "epoch": 25.77319587628866,
      "grad_norm": 0.028440319001674652,
      "learning_rate": 1.0670103092783506e-05,
      "loss": 0.0091,
      "step": 5000
    },
    {
      "epoch": 26.0,
      "eval_f1": 0.9438860971524288,
      "eval_loss": 0.47594669461250305,
      "eval_runtime": 3.3703,
      "eval_samples_per_second": 1062.816,
      "eval_steps_per_second": 8.308,
      "step": 5044
    },
    {
      "epoch": 27.0,
      "eval_f1": 0.9463986599664992,
      "eval_loss": 0.45179277658462524,
      "eval_runtime": 3.3499,
      "eval_samples_per_second": 1069.276,
      "eval_steps_per_second": 8.358,
      "step": 5238
    },
    {
      "epoch": 28.0,
      "eval_f1": 0.9461194863204914,
      "eval_loss": 0.4447253346443176,
      "eval_runtime": 3.3396,
      "eval_samples_per_second": 1072.582,
      "eval_steps_per_second": 8.384,
      "step": 5432
    },
    {
      "epoch": 28.350515463917525,
      "grad_norm": 0.09605555236339569,
      "learning_rate": 8.737113402061856e-06,
      "loss": 0.0085,
      "step": 5500
    },
    {
      "epoch": 29.0,
      "eval_f1": 0.9455611390284757,
      "eval_loss": 0.4533209502696991,
      "eval_runtime": 3.3431,
      "eval_samples_per_second": 1071.445,
      "eval_steps_per_second": 8.375,
      "step": 5626
    },
    {
      "epoch": 30.0,
      "eval_f1": 0.9458403126744835,
      "eval_loss": 0.4549112319946289,
      "eval_runtime": 3.339,
      "eval_samples_per_second": 1072.783,
      "eval_steps_per_second": 8.386,
      "step": 5820
    },
    {
      "epoch": 30.927835051546392,
      "grad_norm": 0.589557409286499,
      "learning_rate": 6.8041237113402065e-06,
      "loss": 0.0071,
      "step": 6000
    },
    {
      "epoch": 31.0,
      "eval_f1": 0.9469570072585148,
      "eval_loss": 0.45267006754875183,
      "eval_runtime": 3.3812,
      "eval_samples_per_second": 1059.377,
      "eval_steps_per_second": 8.281,
      "step": 6014
    },
    {
      "epoch": 32.0,
      "eval_f1": 0.9452819653824679,
      "eval_loss": 0.4533182382583618,
      "eval_runtime": 3.3589,
      "eval_samples_per_second": 1066.414,
      "eval_steps_per_second": 8.336,
      "step": 6208
    },
    {
      "epoch": 33.0,
      "eval_f1": 0.948073701842546,
      "eval_loss": 0.4553990066051483,
      "eval_runtime": 3.3534,
      "eval_samples_per_second": 1068.168,
      "eval_steps_per_second": 8.35,
      "step": 6402
    },
    {
      "epoch": 33.50515463917526,
      "grad_norm": 0.004464476834982634,
      "learning_rate": 4.871134020618557e-06,
      "loss": 0.0072,
      "step": 6500
    },
    {
      "epoch": 34.0,
      "eval_f1": 0.9447236180904522,
      "eval_loss": 0.462903767824173,
      "eval_runtime": 3.3763,
      "eval_samples_per_second": 1060.91,
      "eval_steps_per_second": 8.293,
      "step": 6596
    },
    {
      "epoch": 35.0,
      "eval_f1": 0.9450027917364601,
      "eval_loss": 0.47262680530548096,
      "eval_runtime": 3.3644,
      "eval_samples_per_second": 1064.678,
      "eval_steps_per_second": 8.322,
      "step": 6790
    },
    {
      "epoch": 36.0,
      "eval_f1": 0.9461194863204914,
      "eval_loss": 0.4510941803455353,
      "eval_runtime": 3.3674,
      "eval_samples_per_second": 1063.734,
      "eval_steps_per_second": 8.315,
      "step": 6984
    },
    {
      "epoch": 36.08247422680412,
      "grad_norm": 0.03277302160859108,
      "learning_rate": 2.938144329896907e-06,
      "loss": 0.0067,
      "step": 7000
    },
    {
      "epoch": 37.0,
      "eval_f1": 0.9447236180904522,
      "eval_loss": 0.4571346044540405,
      "eval_runtime": 3.3879,
      "eval_samples_per_second": 1057.285,
      "eval_steps_per_second": 8.265,
      "step": 7178
    },
    {
      "epoch": 38.0,
      "eval_f1": 0.9455611390284757,
      "eval_loss": 0.46757060289382935,
      "eval_runtime": 3.396,
      "eval_samples_per_second": 1054.781,
      "eval_steps_per_second": 8.245,
      "step": 7372
    },
    {
      "epoch": 38.65979381443299,
      "grad_norm": 0.03082508221268654,
      "learning_rate": 1.0051546391752577e-06,
      "loss": 0.0062,
      "step": 7500
    },
    {
      "epoch": 39.0,
      "eval_f1": 0.9461194863204914,
      "eval_loss": 0.4639934301376343,
      "eval_runtime": 3.38,
      "eval_samples_per_second": 1059.749,
      "eval_steps_per_second": 8.284,
      "step": 7566
    },
    {
      "epoch": 40.0,
      "eval_f1": 0.9463986599664992,
      "eval_loss": 0.46385136246681213,
      "eval_runtime": 3.3218,
      "eval_samples_per_second": 1078.319,
      "eval_steps_per_second": 8.429,
      "step": 7760
    },
    {
      "epoch": 40.0,
      "step": 7760,
      "total_flos": 6.517582470540288e+16,
      "train_loss": 0.02645080839235758,
      "train_runtime": 2544.3454,
      "train_samples_per_second": 389.428,
      "train_steps_per_second": 3.05
    }
  ],
  "logging_steps": 500,
  "max_steps": 7760,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "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": 6.517582470540288e+16,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}