{
  "best_metric": 0.05933361500501633,
  "best_model_checkpoint": "runs/roberta-base-500000-samples-512-max-len-64-train-batch-size-8-test-batch-size-3-epochs-1e-05-lr-0.1-warmup-ratio/checkpoint-12000",
  "epoch": 2.1331058020477816,
  "eval_steps": 1500,
  "global_step": 15000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_accuracy": 0.4971,
      "eval_f1": 0.6639222657346396,
      "eval_loss": 0.6927798390388489,
      "eval_precision": 0.4973866025833584,
      "eval_recall": 0.9981112361356695,
      "eval_runtime": 372.4207,
      "eval_samples_per_second": 134.257,
      "eval_steps_per_second": 16.782,
      "step": 0
    },
    {
      "epoch": 0.07110352673492605,
      "grad_norm": 11.443495750427246,
      "learning_rate": 2.369668246445498e-06,
      "loss": 0.4665,
      "step": 500
    },
    {
      "epoch": 0.1422070534698521,
      "grad_norm": 5.395324230194092,
      "learning_rate": 4.739336492890996e-06,
      "loss": 0.1226,
      "step": 1000
    },
    {
      "epoch": 0.21331058020477817,
      "grad_norm": 3.713897705078125,
      "learning_rate": 7.1090047393364935e-06,
      "loss": 0.1013,
      "step": 1500
    },
    {
      "epoch": 0.21331058020477817,
      "eval_accuracy": 0.96946,
      "eval_f1": 0.9694067677759302,
      "eval_loss": 0.08184666186571121,
      "eval_precision": 0.9665987454552719,
      "eval_recall": 0.9722311525478219,
      "eval_runtime": 369.1947,
      "eval_samples_per_second": 135.43,
      "eval_steps_per_second": 16.929,
      "step": 1500
    },
    {
      "epoch": 0.2844141069397042,
      "grad_norm": 22.85436248779297,
      "learning_rate": 9.478672985781992e-06,
      "loss": 0.0911,
      "step": 2000
    },
    {
      "epoch": 0.35551763367463024,
      "grad_norm": 2.183819532394409,
      "learning_rate": 9.794585484040872e-06,
      "loss": 0.0904,
      "step": 2500
    },
    {
      "epoch": 0.42662116040955633,
      "grad_norm": 3.5085792541503906,
      "learning_rate": 9.53123354050353e-06,
      "loss": 0.0853,
      "step": 3000
    },
    {
      "epoch": 0.42662116040955633,
      "eval_accuracy": 0.97452,
      "eval_f1": 0.97444332998997,
      "eval_loss": 0.06914982199668884,
      "eval_precision": 0.9728430665705359,
      "eval_recall": 0.9760488667416815,
      "eval_runtime": 369.1227,
      "eval_samples_per_second": 135.456,
      "eval_steps_per_second": 16.932,
      "step": 3000
    },
    {
      "epoch": 0.49772468714448237,
      "grad_norm": 6.5320234298706055,
      "learning_rate": 9.267881596966186e-06,
      "loss": 0.0799,
      "step": 3500
    },
    {
      "epoch": 0.5688282138794084,
      "grad_norm": 1.3349543809890747,
      "learning_rate": 9.004529653428843e-06,
      "loss": 0.0751,
      "step": 4000
    },
    {
      "epoch": 0.6399317406143344,
      "grad_norm": 1.4865925312042236,
      "learning_rate": 8.7411777098915e-06,
      "loss": 0.0742,
      "step": 4500
    },
    {
      "epoch": 0.6399317406143344,
      "eval_accuracy": 0.97572,
      "eval_f1": 0.9756136756257282,
      "eval_loss": 0.06528624147176743,
      "eval_precision": 0.9753393846895333,
      "eval_recall": 0.9758881208808873,
      "eval_runtime": 369.1379,
      "eval_samples_per_second": 135.451,
      "eval_steps_per_second": 16.931,
      "step": 4500
    },
    {
      "epoch": 0.7110352673492605,
      "grad_norm": 1.9182387590408325,
      "learning_rate": 8.477825766354156e-06,
      "loss": 0.0761,
      "step": 5000
    },
    {
      "epoch": 0.7821387940841866,
      "grad_norm": 4.340898036956787,
      "learning_rate": 8.214473822816812e-06,
      "loss": 0.0742,
      "step": 5500
    },
    {
      "epoch": 0.8532423208191127,
      "grad_norm": 4.503939151763916,
      "learning_rate": 7.95112187927947e-06,
      "loss": 0.0722,
      "step": 6000
    },
    {
      "epoch": 0.8532423208191127,
      "eval_accuracy": 0.97488,
      "eval_f1": 0.9744964262508122,
      "eval_loss": 0.06521258503198624,
      "eval_precision": 0.9848957478246594,
      "eval_recall": 0.9643144189037133,
      "eval_runtime": 369.1106,
      "eval_samples_per_second": 135.461,
      "eval_steps_per_second": 16.933,
      "step": 6000
    },
    {
      "epoch": 0.9243458475540387,
      "grad_norm": 5.767714977264404,
      "learning_rate": 7.687769935742126e-06,
      "loss": 0.0664,
      "step": 6500
    },
    {
      "epoch": 0.9954493742889647,
      "grad_norm": 1.0236719846725464,
      "learning_rate": 7.424417992204783e-06,
      "loss": 0.0687,
      "step": 7000
    },
    {
      "epoch": 1.0665529010238908,
      "grad_norm": 5.575131416320801,
      "learning_rate": 7.1610660486674395e-06,
      "loss": 0.0594,
      "step": 7500
    },
    {
      "epoch": 1.0665529010238908,
      "eval_accuracy": 0.9734,
      "eval_f1": 0.9729013854930725,
      "eval_loss": 0.07639238238334656,
      "eval_precision": 0.9867333443544387,
      "eval_recall": 0.9594518566146921,
      "eval_runtime": 369.1258,
      "eval_samples_per_second": 135.455,
      "eval_steps_per_second": 16.932,
      "step": 7500
    },
    {
      "epoch": 1.1376564277588168,
      "grad_norm": 1.4480912685394287,
      "learning_rate": 6.8977141051300965e-06,
      "loss": 0.0613,
      "step": 8000
    },
    {
      "epoch": 1.2087599544937428,
      "grad_norm": 14.01652717590332,
      "learning_rate": 6.6343621615927535e-06,
      "loss": 0.0626,
      "step": 8500
    },
    {
      "epoch": 1.2798634812286689,
      "grad_norm": 2.645029067993164,
      "learning_rate": 6.3710102180554104e-06,
      "loss": 0.0595,
      "step": 9000
    },
    {
      "epoch": 1.2798634812286689,
      "eval_accuracy": 0.97674,
      "eval_f1": 0.9764427069618586,
      "eval_loss": 0.0677267462015152,
      "eval_precision": 0.9843986113947315,
      "eval_recall": 0.968614370679955,
      "eval_runtime": 369.1238,
      "eval_samples_per_second": 135.456,
      "eval_steps_per_second": 16.932,
      "step": 9000
    },
    {
      "epoch": 1.350967007963595,
      "grad_norm": 4.5592122077941895,
      "learning_rate": 6.1076582745180666e-06,
      "loss": 0.0618,
      "step": 9500
    },
    {
      "epoch": 1.4220705346985212,
      "grad_norm": 5.417106628417969,
      "learning_rate": 5.8443063309807235e-06,
      "loss": 0.058,
      "step": 10000
    },
    {
      "epoch": 1.493174061433447,
      "grad_norm": 1.136661171913147,
      "learning_rate": 5.5809543874433805e-06,
      "loss": 0.0542,
      "step": 10500
    },
    {
      "epoch": 1.493174061433447,
      "eval_accuracy": 0.97848,
      "eval_f1": 0.9783422567529487,
      "eval_loss": 0.06500901281833649,
      "eval_precision": 0.9800387127994193,
      "eval_recall": 0.9766516637196592,
      "eval_runtime": 369.2146,
      "eval_samples_per_second": 135.423,
      "eval_steps_per_second": 16.928,
      "step": 10500
    },
    {
      "epoch": 1.5642775881683733,
      "grad_norm": 2.5331344604492188,
      "learning_rate": 5.317602443906037e-06,
      "loss": 0.0623,
      "step": 11000
    },
    {
      "epoch": 1.635381114903299,
      "grad_norm": 2.5099124908447266,
      "learning_rate": 5.054250500368693e-06,
      "loss": 0.0617,
      "step": 11500
    },
    {
      "epoch": 1.7064846416382253,
      "grad_norm": 0.18802767992019653,
      "learning_rate": 4.79089855683135e-06,
      "loss": 0.0571,
      "step": 12000
    },
    {
      "epoch": 1.7064846416382253,
      "eval_accuracy": 0.97944,
      "eval_f1": 0.9793847511330366,
      "eval_loss": 0.05933361500501633,
      "eval_precision": 0.9774637739172204,
      "eval_recall": 0.9813132936826877,
      "eval_runtime": 369.1545,
      "eval_samples_per_second": 135.445,
      "eval_steps_per_second": 16.931,
      "step": 12000
    },
    {
      "epoch": 1.7775881683731511,
      "grad_norm": 0.17306402325630188,
      "learning_rate": 4.527546613294007e-06,
      "loss": 0.0575,
      "step": 12500
    },
    {
      "epoch": 1.8486916951080774,
      "grad_norm": 2.0170910358428955,
      "learning_rate": 4.264194669756664e-06,
      "loss": 0.0573,
      "step": 13000
    },
    {
      "epoch": 1.9197952218430034,
      "grad_norm": 1.0754927396774292,
      "learning_rate": 4.00084272621932e-06,
      "loss": 0.0562,
      "step": 13500
    },
    {
      "epoch": 1.9197952218430034,
      "eval_accuracy": 0.9793,
      "eval_f1": 0.9792272955343703,
      "eval_loss": 0.05992409214377403,
      "eval_precision": 0.9781083356721864,
      "eval_recall": 0.9803488185179232,
      "eval_runtime": 369.2584,
      "eval_samples_per_second": 135.407,
      "eval_steps_per_second": 16.926,
      "step": 13500
    },
    {
      "epoch": 1.9908987485779295,
      "grad_norm": 0.5176452398300171,
      "learning_rate": 3.7374907826819767e-06,
      "loss": 0.0553,
      "step": 14000
    },
    {
      "epoch": 2.0620022753128557,
      "grad_norm": 3.9174857139587402,
      "learning_rate": 3.474138839144633e-06,
      "loss": 0.0506,
      "step": 14500
    },
    {
      "epoch": 2.1331058020477816,
      "grad_norm": 2.6643998622894287,
      "learning_rate": 3.21078689560729e-06,
      "loss": 0.0463,
      "step": 15000
    },
    {
      "epoch": 2.1331058020477816,
      "eval_accuracy": 0.97976,
      "eval_f1": 0.9796746334605343,
      "eval_loss": 0.05948900803923607,
      "eval_precision": 0.9792419497309885,
      "eval_recall": 0.9801076997267321,
      "eval_runtime": 369.2026,
      "eval_samples_per_second": 135.427,
      "eval_steps_per_second": 16.928,
      "step": 15000
    }
  ],
  "logging_steps": 500,
  "max_steps": 21096,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.5256135448428544e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}