run-20/checkpoint-766/trainer_state.json

{
  "best_metric": 0.8043052837573386,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-766",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 766,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.1449050903320312,
      "learning_rate": 0.00010427761773013143,
      "loss": 0.5295,
      "step": 766
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8043052837573386,
      "eval_f1": 0.82174688057041,
      "eval_loss": 0.44088178873062134,
      "eval_precision": 0.7545008183306056,
      "eval_recall": 0.9021526418786693,
      "eval_runtime": 133.3851,
      "eval_samples_per_second": 7.662,
      "eval_steps_per_second": 1.919,
      "step": 766
    }
  ],
  "logging_steps": 500,
  "max_steps": 3830,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 235695197280.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8705195866732912,
    "learning_rate": 0.00013034702216266429,
    "num_train_epochs": 5,
    "temperature": 7
  }
}