run-16/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7322788517867604,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-16/checkpoint-428",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 428,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.7153942584991455,
      "learning_rate": 0.0004713975088977992,
      "loss": 0.4261,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7041593438781488,
      "eval_f1": 0.3711083437110834,
      "eval_loss": 0.4133966565132141,
      "eval_mcc": 0.25654012586617897,
      "eval_precision": 0.6367521367521367,
      "eval_recall": 0.2618629173989455,
      "eval_runtime": 3.146,
      "eval_samples_per_second": 542.601,
      "eval_steps_per_second": 17.165,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4420157670974731,
      "learning_rate": 0.00040405500762668504,
      "loss": 0.3984,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7322788517867604,
      "eval_f1": 0.4148527528809219,
      "eval_loss": 0.3995412588119507,
      "eval_mcc": 0.3441494170095203,
      "eval_precision": 0.7641509433962265,
      "eval_recall": 0.28471001757469244,
      "eval_runtime": 3.1311,
      "eval_samples_per_second": 545.175,
      "eval_steps_per_second": 17.246,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 1712,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6774741209148297,
    "learning_rate": 0.0005387400101689134,
    "num_train_epochs": 8,
    "temperature": 18
  }
}