trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.67,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 1.5642,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8655619345859429,
      "eval_f1": 0.5773226042428675,
      "eval_loss": 1.0524519681930542,
      "eval_precision": 0.6683604336043361,
      "eval_recall": 0.5081122843162503,
      "eval_runtime": 8.3021,
      "eval_samples_per_second": 287.76,
      "eval_steps_per_second": 36.015,
      "step": 750
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 1.0845,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.2e-05,
      "loss": 0.7686,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9002479123173278,
      "eval_f1": 0.6962228770425931,
      "eval_loss": 0.6765145659446716,
      "eval_precision": 0.7253698018420318,
      "eval_recall": 0.6693278392995107,
      "eval_runtime": 8.4509,
      "eval_samples_per_second": 282.691,
      "eval_steps_per_second": 35.381,
      "step": 1500
    },
    {
      "epoch": 2.67,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.6293,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9122738343771747,
      "eval_f1": 0.7344603381014305,
      "eval_loss": 0.5619152188301086,
      "eval_precision": 0.7417914368268979,
      "eval_recall": 0.7272727272727273,
      "eval_runtime": 8.4191,
      "eval_samples_per_second": 283.761,
      "eval_steps_per_second": 35.515,
      "step": 2250
    }
  ],
  "logging_steps": 500,
  "max_steps": 3750,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 420955855136784.0,
  "trial_name": null,
  "trial_params": null
}