trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.84,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 1.2555780410766602,
      "learning_rate": 0.00018333333333333334,
      "loss": 9.4154,
      "step": 2
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.53036367893219,
      "learning_rate": 0.0001666666666666667,
      "loss": 9.0215,
      "step": 4
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8442403078079224,
      "learning_rate": 0.00015000000000000001,
      "loss": 8.535,
      "step": 6
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.2389612197875977,
      "learning_rate": 0.00013333333333333334,
      "loss": 8.0172,
      "step": 8
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.622530698776245,
      "learning_rate": 0.00011666666666666668,
      "loss": 7.8423,
      "step": 10
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.8883254528045654,
      "learning_rate": 0.0001,
      "loss": 7.6953,
      "step": 12
    },
    {
      "epoch": 2.24,
      "grad_norm": NaN,
      "learning_rate": 9.166666666666667e-05,
      "loss": 7.4797,
      "step": 14
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.4147595167160034,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.4009,
      "step": 16
    },
    {
      "epoch": 2.88,
      "grad_norm": 2.086115598678589,
      "learning_rate": 5.833333333333334e-05,
      "loss": 7.2106,
      "step": 18
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.3016234636306763,
      "learning_rate": 4.166666666666667e-05,
      "loss": 7.2342,
      "step": 20
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.3679742813110352,
      "learning_rate": 2.5e-05,
      "loss": 7.1718,
      "step": 22
    },
    {
      "epoch": 3.84,
      "grad_norm": 2.1608622074127197,
      "learning_rate": 8.333333333333334e-06,
      "loss": 7.1113,
      "step": 24
    },
    {
      "epoch": 3.84,
      "step": 24,
      "total_flos": 124647858838128.0,
      "train_loss": 7.844584743181865,
      "train_runtime": 146.2747,
      "train_samples_per_second": 2.735,
      "train_steps_per_second": 0.164
    }
  ],
  "logging_steps": 2,
  "max_steps": 24,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 124647858838128.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}