trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.810126582278481,
  "eval_steps": 500,
  "global_step": 4,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.9072078083746264,
      "learning_rate": 2e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.766769690439105,
      "num_tokens": 524288.0,
      "step": 1
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 4.753814194553616,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.8217,
      "mean_token_accuracy": 0.7695875316858292,
      "num_tokens": 1048576.0,
      "step": 2
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 78.58612697641571,
      "learning_rate": 1e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7892876267433167,
      "num_tokens": 1572864.0,
      "step": 3
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.129716830110077,
      "learning_rate": 5e-06,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7917388044297695,
      "num_tokens": 2094002.0,
      "step": 4
    },
    {
      "epoch": 0.810126582278481,
      "step": 4,
      "total_flos": 4154970079232.0,
      "train_loss": 0.8020447641611099,
      "train_runtime": 203.4558,
      "train_samples_per_second": 3.092,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 4,
  "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": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4154970079232.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}