{
  "best_metric": 0.9664750957854407,
  "best_model_checkpoint": "bert-finetuned-spam/run-0/checkpoint-835",
  "epoch": 4.982035928143713,
  "eval_steps": 500,
  "global_step": 1040,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9964071856287425,
      "eval_accuracy": 0.9396551724137931,
      "eval_loss": 0.1946447342634201,
      "eval_runtime": 12.0187,
      "eval_samples_per_second": 86.865,
      "eval_steps_per_second": 5.491,
      "step": 208
    },
    {
      "epoch": 1.9976047904191616,
      "eval_accuracy": 0.9645593869731801,
      "eval_loss": 0.10755528509616852,
      "eval_runtime": 12.0129,
      "eval_samples_per_second": 86.907,
      "eval_steps_per_second": 5.494,
      "step": 417
    },
    {
      "epoch": 2.3952095808383236,
      "grad_norm": 1.271166205406189,
      "learning_rate": 2.8464605725436264e-05,
      "loss": 0.3008,
      "step": 500
    },
    {
      "epoch": 2.998802395209581,
      "eval_accuracy": 0.9655172413793104,
      "eval_loss": 0.09413430839776993,
      "eval_runtime": 12.0301,
      "eval_samples_per_second": 86.782,
      "eval_steps_per_second": 5.486,
      "step": 626
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9664750957854407,
      "eval_loss": 0.10097901523113251,
      "eval_runtime": 12.0003,
      "eval_samples_per_second": 86.998,
      "eval_steps_per_second": 5.5,
      "step": 835
    },
    {
      "epoch": 4.790419161676647,
      "grad_norm": 0.1700475811958313,
      "learning_rate": 2.108489312995279e-06,
      "loss": 0.1168,
      "step": 1000
    },
    {
      "epoch": 4.982035928143713,
      "eval_accuracy": 0.9664750957854407,
      "eval_loss": 0.10203641653060913,
      "eval_runtime": 11.9796,
      "eval_samples_per_second": 87.148,
      "eval_steps_per_second": 5.509,
      "step": 1040
    }
  ],
  "logging_steps": 500,
  "max_steps": 1040,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 3649464194725728.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 5.482072213787725e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 2
  }
}