{
  "best_metric": 0.6162846088409424,
  "best_model_checkpoint": "BERT-WMM/run-13/checkpoint-533",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1066,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.94,
      "grad_norm": 12.600181579589844,
      "learning_rate": 1.4607577277616791e-05,
      "loss": 0.7148,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7577464788732394,
      "eval_loss": 0.6162846088409424,
      "eval_runtime": 1.9944,
      "eval_samples_per_second": 1067.994,
      "eval_steps_per_second": 67.188,
      "step": 533
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.747657775878906,
      "learning_rate": 1.1234002848605984e-05,
      "loss": 0.4485,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7624413145539906,
      "eval_loss": 0.6197112202644348,
      "eval_runtime": 2.1341,
      "eval_samples_per_second": 998.1,
      "eval_steps_per_second": 62.791,
      "step": 1066
    }
  ],
  "logging_steps": 500,
  "max_steps": 2665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 338261076519408.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.7981151706627596e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 16,
    "weight_decay": 0.009594844383188954
  }
}