{
  "best_metric": 0.715109646320343,
  "best_model_checkpoint": "BERT-WMM/run-4/checkpoint-1599",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1599,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.94,
      "grad_norm": 11.732254028320312,
      "learning_rate": 9.02435978576971e-07,
      "loss": 0.9819,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6502347417840375,
      "eval_loss": 0.8438959121704102,
      "eval_runtime": 2.0462,
      "eval_samples_per_second": 1040.962,
      "eval_steps_per_second": 65.488,
      "step": 533
    },
    {
      "epoch": 1.88,
      "grad_norm": 25.85025978088379,
      "learning_rate": 4.918645597521434e-07,
      "loss": 0.8112,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7023474178403756,
      "eval_loss": 0.7426179647445679,
      "eval_runtime": 2.1113,
      "eval_samples_per_second": 1008.876,
      "eval_steps_per_second": 63.469,
      "step": 1066
    },
    {
      "epoch": 2.81,
      "grad_norm": 10.392358779907227,
      "learning_rate": 8.129314092731586e-08,
      "loss": 0.7317,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7145539906103286,
      "eval_loss": 0.715109646320343,
      "eval_runtime": 2.0834,
      "eval_samples_per_second": 1022.379,
      "eval_steps_per_second": 64.319,
      "step": 1599
    }
  ],
  "logging_steps": 500,
  "max_steps": 1599,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 507646505902536.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.3130073974017986e-06
  }
}