{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.996539792387544,
  "global_step": 3600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8460490703582764,
      "eval_loss": 0.5365713834762573,
      "eval_runtime": 8.0333,
      "eval_samples_per_second": 91.369,
      "eval_steps_per_second": 45.685,
      "step": 180
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8092643022537231,
      "eval_loss": 0.5189609527587891,
      "eval_runtime": 8.0124,
      "eval_samples_per_second": 91.608,
      "eval_steps_per_second": 45.804,
      "step": 360
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.309722222222222e-05,
      "loss": 0.4021,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8283378481864929,
      "eval_loss": 0.6708077788352966,
      "eval_runtime": 8.0551,
      "eval_samples_per_second": 91.122,
      "eval_steps_per_second": 45.561,
      "step": 540
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8542234301567078,
      "eval_loss": 0.516476571559906,
      "eval_runtime": 8.0611,
      "eval_samples_per_second": 91.054,
      "eval_steps_per_second": 45.527,
      "step": 720
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8188011050224304,
      "eval_loss": 0.6029361486434937,
      "eval_runtime": 8.0681,
      "eval_samples_per_second": 90.975,
      "eval_steps_per_second": 45.488,
      "step": 900
    },
    {
      "epoch": 5.55,
      "learning_rate": 3.6166666666666674e-05,
      "loss": 0.2576,
      "step": 1000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8487738370895386,
      "eval_loss": 0.6060934066772461,
      "eval_runtime": 8.0522,
      "eval_samples_per_second": 91.155,
      "eval_steps_per_second": 45.578,
      "step": 1080
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8514986634254456,
      "eval_loss": 0.748849630355835,
      "eval_runtime": 8.0692,
      "eval_samples_per_second": 90.963,
      "eval_steps_per_second": 45.481,
      "step": 1260
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8651226162910461,
      "eval_loss": 0.6119422912597656,
      "eval_runtime": 8.0635,
      "eval_samples_per_second": 91.028,
      "eval_steps_per_second": 45.514,
      "step": 1440
    },
    {
      "epoch": 8.33,
      "learning_rate": 2.9236111111111115e-05,
      "loss": 0.1738,
      "step": 1500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8542234301567078,
      "eval_loss": 0.6864181160926819,
      "eval_runtime": 8.212,
      "eval_samples_per_second": 89.382,
      "eval_steps_per_second": 44.691,
      "step": 1620
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8446866273880005,
      "eval_loss": 0.7817405462265015,
      "eval_runtime": 8.0215,
      "eval_samples_per_second": 91.505,
      "eval_steps_per_second": 45.752,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8514986634254456,
      "eval_loss": 0.6188392043113708,
      "eval_runtime": 8.0857,
      "eval_samples_per_second": 90.777,
      "eval_steps_per_second": 45.389,
      "step": 1980
    },
    {
      "epoch": 11.11,
      "learning_rate": 2.2305555555555556e-05,
      "loss": 0.1303,
      "step": 2000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8569482564926147,
      "eval_loss": 0.5936163663864136,
      "eval_runtime": 8.0618,
      "eval_samples_per_second": 91.047,
      "eval_steps_per_second": 45.523,
      "step": 2160
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.859673023223877,
      "eval_loss": 0.6109394431114197,
      "eval_runtime": 8.0512,
      "eval_samples_per_second": 91.167,
      "eval_steps_per_second": 45.583,
      "step": 2340
    },
    {
      "epoch": 13.89,
      "learning_rate": 1.5375e-05,
      "loss": 0.1226,
      "step": 2500
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8501362204551697,
      "eval_loss": 0.7600889205932617,
      "eval_runtime": 8.0176,
      "eval_samples_per_second": 91.549,
      "eval_steps_per_second": 45.774,
      "step": 2520
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8501362204551697,
      "eval_loss": 0.6596993803977966,
      "eval_runtime": 8.0565,
      "eval_samples_per_second": 91.107,
      "eval_steps_per_second": 45.553,
      "step": 2700
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8460490703582764,
      "eval_loss": 0.712175190448761,
      "eval_runtime": 8.0456,
      "eval_samples_per_second": 91.23,
      "eval_steps_per_second": 45.615,
      "step": 2880
    },
    {
      "epoch": 16.66,
      "learning_rate": 8.430555555555556e-06,
      "loss": 0.1261,
      "step": 3000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8514986634254456,
      "eval_loss": 0.7294248938560486,
      "eval_runtime": 8.0369,
      "eval_samples_per_second": 91.329,
      "eval_steps_per_second": 45.664,
      "step": 3060
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.863760232925415,
      "eval_loss": 0.6875426173210144,
      "eval_runtime": 8.0764,
      "eval_samples_per_second": 90.882,
      "eval_steps_per_second": 45.441,
      "step": 3240
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8664849996566772,
      "eval_loss": 0.6823601126670837,
      "eval_runtime": 8.0997,
      "eval_samples_per_second": 90.62,
      "eval_steps_per_second": 45.31,
      "step": 3420
    },
    {
      "epoch": 19.44,
      "learning_rate": 1.4861111111111113e-06,
      "loss": 0.1044,
      "step": 3500
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8623978495597839,
      "eval_loss": 0.6752045750617981,
      "eval_runtime": 8.0488,
      "eval_samples_per_second": 91.193,
      "eval_steps_per_second": 45.597,
      "step": 3600
    },
    {
      "epoch": 20.0,
      "step": 3600,
      "total_flos": 1.3459697175036672e+16,
      "train_loss": 0.18581116994222005,
      "train_runtime": 2514.2374,
      "train_samples_per_second": 22.981,
      "train_steps_per_second": 1.432
    }
  ],
  "max_steps": 3600,
  "num_train_epochs": 20,
  "total_flos": 1.3459697175036672e+16,
  "trial_name": null,
  "trial_params": null
}