last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.00013455328310010763,
  "eval_steps": 500,
  "global_step": 2,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.727664155005382e-05,
      "grad_norm": 16.59201431274414,
      "learning_rate": 0.0001,
      "loss": 7.986,
      "step": 1
    },
    {
      "epoch": 0.00013455328310010763,
      "grad_norm": 3.2888948917388916,
      "learning_rate": 0.0002,
      "loss": 2.0306,
      "step": 2
    },
    {
      "epoch": 0.00013455328310010763,
      "eval_loss": 2.1215152740478516,
      "eval_runtime": 5.4188,
      "eval_samples_per_second": 27.866,
      "eval_steps_per_second": 27.866,
      "step": 2
    }
  ],
  "logging_steps": 1,
  "max_steps": 2,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1598699864064.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}