trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 65.0,
      "learning_rate": 0.0002,
      "loss": 15.8332,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 16.63365364074707,
      "eval_runtime": 2.5247,
      "eval_samples_per_second": 0.792,
      "eval_steps_per_second": 0.396,
      "step": 1
    },
    {
      "epoch": 1.0,
      "step": 1,
      "total_flos": 6135154623905792.0,
      "train_loss": 15.833246231079102,
      "train_runtime": 7.5505,
      "train_samples_per_second": 7.814,
      "train_steps_per_second": 0.132
    }
  ],
  "logging_steps": 5,
  "max_steps": 1,
  "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
}