trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 132.0,
      "learning_rate": 0.0002,
      "loss": 17.0553,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 7.457533836364746,
      "eval_runtime": 4.1222,
      "eval_samples_per_second": 0.485,
      "eval_steps_per_second": 0.243,
      "step": 4
    },
    {
      "epoch": 1.0,
      "step": 4,
      "total_flos": 6135154623905792.0,
      "train_loss": 12.472952842712402,
      "train_runtime": 44.3036,
      "train_samples_per_second": 1.332,
      "train_steps_per_second": 0.09
    }
  ],
  "logging_steps": 5,
  "max_steps": 4,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 6135154623905792.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}