{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.3748186528682709,
      "learning_rate": 0.0018976410256410258,
      "loss": 0.4066,
      "step": 500
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.34374505281448364,
      "learning_rate": 0.001795076923076923,
      "loss": 0.1902,
      "step": 1000
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.29863688349723816,
      "learning_rate": 0.0016925128205128206,
      "loss": 0.1362,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.0691862627863884,
      "eval_runtime": 22.7429,
      "eval_samples_per_second": 304.799,
      "eval_steps_per_second": 9.541,
      "step": 1950
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 0.25807419419288635,
      "learning_rate": 0.001589948717948718,
      "loss": 0.1005,
      "step": 2000
    },
    {
      "epoch": 1.282051282051282,
      "grad_norm": 0.24320101737976074,
      "learning_rate": 0.0014873846153846152,
      "loss": 0.0615,
      "step": 2500
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.24260133504867554,
      "learning_rate": 0.001384820512820513,
      "loss": 0.0562,
      "step": 3000
    },
    {
      "epoch": 1.7948717948717947,
      "grad_norm": 0.4375219941139221,
      "learning_rate": 0.0012822564102564103,
      "loss": 0.0466,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.03345639631152153,
      "eval_runtime": 23.0664,
      "eval_samples_per_second": 300.523,
      "eval_steps_per_second": 9.408,
      "step": 3900
    }
  ],
  "logging_steps": 500,
  "max_steps": 9750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.87428490268672e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}