{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 2440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.6188464164733887,
      "learning_rate": 4.75e-05,
      "loss": 0.5579,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7092731829573935,
      "eval_f1": 0.667786391042205,
      "eval_loss": 0.5389899015426636,
      "eval_precision": 0.6626257628236847,
      "eval_recall": 0.6793053282414985,
      "eval_runtime": 1.7107,
      "eval_samples_per_second": 233.243,
      "eval_steps_per_second": 29.228,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.044891595840454,
      "learning_rate": 4.5e-05,
      "loss": 0.5043,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7644110275689223,
      "eval_f1": 0.6425003812719231,
      "eval_loss": 0.48353031277656555,
      "eval_precision": 0.7516326530612245,
      "eval_recall": 0.6307965084560829,
      "eval_runtime": 1.7072,
      "eval_samples_per_second": 233.71,
      "eval_steps_per_second": 29.287,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.0588178634643555,
      "learning_rate": 4.25e-05,
      "loss": 0.4819,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7769423558897243,
      "eval_f1": 0.7150114361381967,
      "eval_loss": 0.45847204327583313,
      "eval_precision": 0.7321575907590759,
      "eval_recall": 0.7046735770140026,
      "eval_runtime": 1.7067,
      "eval_samples_per_second": 233.787,
      "eval_steps_per_second": 29.297,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.7940452098846436,
      "learning_rate": 4e-05,
      "loss": 0.4474,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7819548872180451,
      "eval_f1": 0.7471685470185073,
      "eval_loss": 0.45870429277420044,
      "eval_precision": 0.7399324134046747,
      "eval_recall": 0.7582287688670667,
      "eval_runtime": 1.7057,
      "eval_samples_per_second": 233.927,
      "eval_steps_per_second": 29.314,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.552182197570801,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4336,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8070175438596491,
      "eval_f1": 0.7504244482173175,
      "eval_loss": 0.42432549595832825,
      "eval_precision": 0.7756337629230987,
      "eval_recall": 0.7359519912711402,
      "eval_runtime": 1.7092,
      "eval_samples_per_second": 233.446,
      "eval_steps_per_second": 29.254,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.7669365406036377,
      "learning_rate": 3.5e-05,
      "loss": 0.4036,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8220551378446115,
      "eval_f1": 0.7889628045918786,
      "eval_loss": 0.39900773763656616,
      "eval_precision": 0.7845601173020528,
      "eval_recall": 0.7940989270776504,
      "eval_runtime": 1.7065,
      "eval_samples_per_second": 233.816,
      "eval_steps_per_second": 29.3,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.482671022415161,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.3871,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8345864661654135,
      "eval_f1": 0.7917273014868713,
      "eval_loss": 0.38430532813072205,
      "eval_precision": 0.8074456774536514,
      "eval_recall": 0.780460083651573,
      "eval_runtime": 1.7068,
      "eval_samples_per_second": 233.771,
      "eval_steps_per_second": 29.295,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.953066825866699,
      "learning_rate": 3e-05,
      "loss": 0.3704,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8370927318295739,
      "eval_f1": 0.7838695967399185,
      "eval_loss": 0.37806496024131775,
      "eval_precision": 0.8269841269841269,
      "eval_recall": 0.7622294962720495,
      "eval_runtime": 1.7054,
      "eval_samples_per_second": 233.968,
      "eval_steps_per_second": 29.319,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 6.039283752441406,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.3563,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8446115288220551,
      "eval_f1": 0.7958506634977223,
      "eval_loss": 0.3728295564651489,
      "eval_precision": 0.8343280282935455,
      "eval_recall": 0.7750500090925623,
      "eval_runtime": 1.7068,
      "eval_samples_per_second": 233.766,
      "eval_steps_per_second": 29.294,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.432938098907471,
      "learning_rate": 2.5e-05,
      "loss": 0.34,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8596491228070176,
      "eval_f1": 0.8262195121951219,
      "eval_loss": 0.354525089263916,
      "eval_precision": 0.8360165151709128,
      "eval_recall": 0.8181942171303873,
      "eval_runtime": 1.7046,
      "eval_samples_per_second": 234.074,
      "eval_steps_per_second": 29.333,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 2.296050548553467,
      "learning_rate": 2.25e-05,
      "loss": 0.3394,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8245369048813042,
      "eval_loss": 0.34459224343299866,
      "eval_precision": 0.8310003145643283,
      "eval_recall": 0.8189216221131115,
      "eval_runtime": 1.7032,
      "eval_samples_per_second": 234.258,
      "eval_steps_per_second": 29.356,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 3.527177572250366,
      "learning_rate": 2e-05,
      "loss": 0.3182,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8596491228070176,
      "eval_f1": 0.8242843661528783,
      "eval_loss": 0.3410705029964447,
      "eval_precision": 0.8389366308055628,
      "eval_recall": 0.8131933078741589,
      "eval_runtime": 1.7024,
      "eval_samples_per_second": 234.371,
      "eval_steps_per_second": 29.37,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 3.337897777557373,
      "learning_rate": 1.75e-05,
      "loss": 0.3226,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8546365914786967,
      "eval_f1": 0.8237632508833923,
      "eval_loss": 0.33531269431114197,
      "eval_precision": 0.8254439681567667,
      "eval_recall": 0.8221494817239499,
      "eval_runtime": 1.7031,
      "eval_samples_per_second": 234.283,
      "eval_steps_per_second": 29.359,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 5.108224868774414,
      "learning_rate": 1.5e-05,
      "loss": 0.3181,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8546365914786967,
      "eval_f1": 0.8271953405017921,
      "eval_loss": 0.33688613772392273,
      "eval_precision": 0.8228172499116295,
      "eval_recall": 0.8321513002364066,
      "eval_runtime": 1.7034,
      "eval_samples_per_second": 234.241,
      "eval_steps_per_second": 29.354,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 4.430032253265381,
      "learning_rate": 1.25e-05,
      "loss": 0.3044,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8263588263588264,
      "eval_loss": 0.3312358856201172,
      "eval_precision": 0.8289473684210527,
      "eval_recall": 0.8239225313693399,
      "eval_runtime": 1.7023,
      "eval_samples_per_second": 234.391,
      "eval_steps_per_second": 29.372,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 4.105480670928955,
      "learning_rate": 1e-05,
      "loss": 0.3038,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8281017346283209,
      "eval_loss": 0.3286840319633484,
      "eval_precision": 0.8272965800108572,
      "eval_recall": 0.8289234406255683,
      "eval_runtime": 1.7022,
      "eval_samples_per_second": 234.402,
      "eval_steps_per_second": 29.374,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 6.878963470458984,
      "learning_rate": 7.5e-06,
      "loss": 0.3033,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8596491228070176,
      "eval_f1": 0.8323529411764705,
      "eval_loss": 0.32676100730895996,
      "eval_precision": 0.8292704679231822,
      "eval_recall": 0.8356973995271868,
      "eval_runtime": 1.7037,
      "eval_samples_per_second": 234.194,
      "eval_steps_per_second": 29.348,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 2.2083144187927246,
      "learning_rate": 5e-06,
      "loss": 0.3018,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8289446964056049,
      "eval_loss": 0.3250683546066284,
      "eval_precision": 0.8266129032258065,
      "eval_recall": 0.8314238952536825,
      "eval_runtime": 1.7036,
      "eval_samples_per_second": 234.208,
      "eval_steps_per_second": 29.349,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 4.615198612213135,
      "learning_rate": 2.5e-06,
      "loss": 0.2955,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8281017346283209,
      "eval_loss": 0.3252684772014618,
      "eval_precision": 0.8272965800108572,
      "eval_recall": 0.8289234406255683,
      "eval_runtime": 1.7036,
      "eval_samples_per_second": 234.205,
      "eval_steps_per_second": 29.349,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 6.056838035583496,
      "learning_rate": 0.0,
      "loss": 0.2999,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.849624060150376,
      "eval_f1": 0.8167483159828537,
      "eval_loss": 0.32519420981407166,
      "eval_precision": 0.8201621387462095,
      "eval_recall": 0.8136024731769412,
      "eval_runtime": 1.7026,
      "eval_samples_per_second": 234.342,
      "eval_steps_per_second": 29.366,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7618923371160000.0,
      "train_loss": 0.3694801080422323,
      "train_runtime": 613.6291,
      "train_samples_per_second": 118.801,
      "train_steps_per_second": 3.976
    }
  ],
  "logging_steps": 500,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7618923371160000.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}