trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 2,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.07062597024649106,
      "learning_rate": 5e-05,
      "loss": 0.3298,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.955711603164673,
      "eval_runtime": 0.7827,
      "eval_samples_per_second": 1.278,
      "eval_steps_per_second": 1.278,
      "step": 1
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.09562765035626467,
      "learning_rate": 0.0,
      "loss": 0.4377,
      "step": 2
    },
    {
      "epoch": 2.0,
      "eval_loss": 3.8781862258911133,
      "eval_runtime": 1.6336,
      "eval_samples_per_second": 0.612,
      "eval_steps_per_second": 0.612,
      "step": 2
    },
    {
      "epoch": 2.0,
      "step": 2,
      "total_flos": 5408059555840.0,
      "train_loss": 0.3837674856185913,
      "train_runtime": 21.2768,
      "train_samples_per_second": 0.846,
      "train_steps_per_second": 0.094
    }
  ],
  "logging_steps": 1,
  "max_steps": 2,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5408059555840.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}