run-6/checkpoint-93/trainer_state.json

{
  "best_metric": 0.5,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-93",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 93,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.4286664128303528,
      "learning_rate": 1.945409831472016e-05,
      "loss": 0.4806,
      "step": 93
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.4703535735607147,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 28.42,
      "eval_samples_per_second": 35.961,
      "eval_steps_per_second": 1.126,
      "step": 93
    }
  ],
  "logging_steps": 500,
  "max_steps": 837,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 235695197280.0,
  "train_batch_size": 33,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6122687021783514,
    "learning_rate": 2.188586060406018e-05,
    "num_train_epochs": 9,
    "per_device_train_batch_size": 33,
    "temperature": 14
  }
}