{
  "best_metric": 0.6060042381286621,
  "best_model_checkpoint": "BERT-WMM/run-3/checkpoint-534",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1068,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7262910798122065,
      "eval_loss": 0.6716192960739136,
      "eval_runtime": 2.087,
      "eval_samples_per_second": 1020.621,
      "eval_steps_per_second": 64.208,
      "step": 267
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.963554382324219,
      "learning_rate": 2.063992947537006e-05,
      "loss": 0.5881,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7652582159624414,
      "eval_loss": 0.6060042381286621,
      "eval_runtime": 2.1035,
      "eval_samples_per_second": 1012.591,
      "eval_steps_per_second": 63.703,
      "step": 534
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7680751173708921,
      "eval_loss": 0.7311103343963623,
      "eval_runtime": 1.9991,
      "eval_samples_per_second": 1065.459,
      "eval_steps_per_second": 67.029,
      "step": 801
    },
    {
      "epoch": 3.75,
      "grad_norm": 5.572795391082764,
      "learning_rate": 2.470977472403458e-06,
      "loss": 0.2017,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7690140845070422,
      "eval_loss": 0.9167852997779846,
      "eval_runtime": 2.1137,
      "eval_samples_per_second": 1007.72,
      "eval_steps_per_second": 63.396,
      "step": 1068
    }
  ],
  "logging_steps": 500,
  "max_steps": 1068,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 697511614823352.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 3.8808881478336665e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 32,
    "weight_decay": 1.7684940065509674e-05
  }
}