trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 1.0748785734176636,
      "eval_runtime": 4.928,
      "eval_samples_per_second": 32.671,
      "eval_steps_per_second": 0.609,
      "step": 6
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 4.1410076721151645,
      "learning_rate": 5e-06,
      "loss": 1.1301,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.0174283981323242,
      "eval_runtime": 4.8291,
      "eval_samples_per_second": 33.34,
      "eval_steps_per_second": 0.621,
      "step": 12
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.9865013957023621,
      "eval_runtime": 3.813,
      "eval_samples_per_second": 42.224,
      "eval_steps_per_second": 0.787,
      "step": 18
    },
    {
      "epoch": 3.0,
      "step": 18,
      "total_flos": 29941290762240.0,
      "train_loss": 1.0634034474690754,
      "train_runtime": 887.8635,
      "train_samples_per_second": 10.329,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 10,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 29941290762240.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}