last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0004854368932038835,
  "eval_steps": 3,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016181229773462783,
      "grad_norm": 0.15954121947288513,
      "learning_rate": 2e-05,
      "loss": 11.9209,
      "step": 1
    },
    {
      "epoch": 0.00016181229773462783,
      "eval_loss": 11.928426742553711,
      "eval_runtime": 39.4665,
      "eval_samples_per_second": 65.955,
      "eval_steps_per_second": 32.99,
      "step": 1
    },
    {
      "epoch": 0.00032362459546925567,
      "grad_norm": 0.11309516429901123,
      "learning_rate": 4e-05,
      "loss": 11.9281,
      "step": 2
    },
    {
      "epoch": 0.0004854368932038835,
      "grad_norm": 0.12080027163028717,
      "learning_rate": 6e-05,
      "loss": 11.9262,
      "step": 3
    },
    {
      "epoch": 0.0004854368932038835,
      "eval_loss": 11.928329467773438,
      "eval_runtime": 35.764,
      "eval_samples_per_second": 72.783,
      "eval_steps_per_second": 36.405,
      "step": 3
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 406756786176.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}