run-0/checkpoint-102/trainer_state.json

{
  "best_metric": 0.9778991937637329,
  "best_model_checkpoint": "BERT-WMM/run-0/checkpoint-102",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 102,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4572769953051643,
      "eval_loss": 1.057793140411377,
      "eval_runtime": 2.1526,
      "eval_samples_per_second": 989.489,
      "eval_steps_per_second": 62.25,
      "step": 34
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.535680751173709,
      "eval_loss": 0.9990364909172058,
      "eval_runtime": 2.1045,
      "eval_samples_per_second": 1012.106,
      "eval_steps_per_second": 63.672,
      "step": 68
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5384976525821596,
      "eval_loss": 0.9778991937637329,
      "eval_runtime": 2.072,
      "eval_samples_per_second": 1028.013,
      "eval_steps_per_second": 64.673,
      "step": 102
    }
  ],
  "logging_steps": 500,
  "max_steps": 136,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.3245164890212277e-06,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 256,
    "weight_decay": 3.536833872428622e-05
  }
}