{
  "best_metric": 0.6660808435852372,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-15/checkpoint-642",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 642,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.24385033547878265,
      "learning_rate": 0.00027237519894122124,
      "loss": 0.0406,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6654950205038078,
      "eval_f1": 0.017211703958691912,
      "eval_loss": 0.03144706040620804,
      "eval_mcc": 0.01487410293271824,
      "eval_precision": 0.4166666666666667,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1499,
      "eval_samples_per_second": 541.926,
      "eval_steps_per_second": 17.144,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.1403072327375412,
      "learning_rate": 0.00021790015915297697,
      "loss": 0.0328,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6631517281780902,
      "eval_f1": 0.033613445378151266,
      "eval_loss": 0.03111909329891205,
      "eval_mcc": 0.013529273078664125,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.01757469244288225,
      "eval_runtime": 3.1379,
      "eval_samples_per_second": 543.986,
      "eval_steps_per_second": 17.209,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.37626323103904724,
      "learning_rate": 0.00016342511936473273,
      "loss": 0.0321,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6660808435852372,
      "eval_f1": 0.01724137931034483,
      "eval_loss": 0.030778612941503525,
      "eval_mcc": 0.020707884164064556,
      "eval_precision": 0.45454545454545453,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1495,
      "eval_samples_per_second": 541.993,
      "eval_steps_per_second": 17.146,
      "step": 642
    }
  ],
  "logging_steps": 500,
  "max_steps": 1284,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1574326993320.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.037718713604154464,
    "learning_rate": 0.00032685023872946545,
    "num_train_epochs": 6,
    "temperature": 24
  }
}