trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 1e-05,
      "loss": 0.8226,
      "step": 250
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9035,
      "eval_f1": 0.9023761650285413,
      "eval_loss": 0.326994925737381,
      "eval_runtime": 4.5192,
      "eval_samples_per_second": 442.555,
      "eval_steps_per_second": 7.081,
      "step": 250
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0,
      "loss": 0.2533,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.922,
      "eval_f1": 0.9219196681926928,
      "eval_loss": 0.21616625785827637,
      "eval_runtime": 4.4445,
      "eval_samples_per_second": 449.996,
      "eval_steps_per_second": 7.2,
      "step": 500
    },
    {
      "epoch": 2.0,
      "step": 500,
      "total_flos": 720342861696000.0,
      "train_loss": 0.5379159927368165,
      "train_runtime": 263.9893,
      "train_samples_per_second": 121.217,
      "train_steps_per_second": 1.894
    }
  ],
  "logging_steps": 250,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 720342861696000.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}