trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 366,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 3e-05,
      "loss": 1.2172,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 67.06197793131896,
      "eval_f1": 71.45312910642082,
      "eval_runtime": 3.7494,
      "eval_samples_per_second": 2586.292,
      "eval_steps_per_second": 8.268,
      "step": 122
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 0.554,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 69.11415901825306,
      "eval_f1": 72.75764323373028,
      "eval_runtime": 3.4296,
      "eval_samples_per_second": 2827.416,
      "eval_steps_per_second": 9.039,
      "step": 244
    },
    {
      "epoch": 3.0,
      "learning_rate": 3e-05,
      "loss": 0.4881,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_exact_match": 70.60946684541611,
      "eval_f1": 73.64681521828038,
      "eval_runtime": 3.3908,
      "eval_samples_per_second": 2859.83,
      "eval_steps_per_second": 9.142,
      "step": 366
    }
  ],
  "max_steps": 1220,
  "num_train_epochs": 10,
  "total_flos": 43819446829056.0,
  "trial_name": null,
  "trial_params": null
}