trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 825,
  "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": 1.8579,
      "step": 275
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 74.99526963103122,
      "eval_f1": 84.08825089968796,
      "eval_runtime": 6.0093,
      "eval_samples_per_second": 1768.246,
      "eval_steps_per_second": 5.658,
      "step": 275
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 1.0254,
      "step": 550
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 78.6565752128666,
      "eval_f1": 86.84668894090845,
      "eval_runtime": 3.8458,
      "eval_samples_per_second": 2762.984,
      "eval_steps_per_second": 8.841,
      "step": 550
    },
    {
      "epoch": 3.0,
      "learning_rate": 3e-05,
      "loss": 0.8243,
      "step": 825
    },
    {
      "epoch": 3.0,
      "eval_exact_match": 79.5837275307474,
      "eval_f1": 87.57352419732825,
      "eval_runtime": 3.8446,
      "eval_samples_per_second": 2763.898,
      "eval_steps_per_second": 8.844,
      "step": 825
    }
  ],
  "max_steps": 5500,
  "num_train_epochs": 20,
  "total_flos": 99220392509440.0,
  "trial_name": null,
  "trial_params": null
}