{
  "best_metric": 0.5817901492118835,
  "best_model_checkpoint": "BERT-WMM/run-13/checkpoint-534",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 801,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7563380281690141,
      "eval_loss": 0.6019130349159241,
      "eval_runtime": 2.4359,
      "eval_samples_per_second": 874.427,
      "eval_steps_per_second": 55.011,
      "step": 267
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.674468994140625,
      "learning_rate": 5.7987153604787005e-06,
      "loss": 0.6061,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7619718309859155,
      "eval_loss": 0.5817901492118835,
      "eval_runtime": 2.4819,
      "eval_samples_per_second": 858.201,
      "eval_steps_per_second": 53.99,
      "step": 534
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7582159624413145,
      "eval_loss": 0.6465383768081665,
      "eval_runtime": 2.4611,
      "eval_samples_per_second": 865.479,
      "eval_steps_per_second": 54.448,
      "step": 801
    }
  ],
  "logging_steps": 500,
  "max_steps": 801,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 349221394826640.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "per_device_train_batch_size": 32
  }
}