run-31/checkpoint-297/trainer_state.json

{
  "best_metric": 0.5504950495049505,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-31/checkpoint-297",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 297,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.7863609790802002,
      "learning_rate": 0.00019509013723494322,
      "loss": 0.6459,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5504950495049505,
      "eval_f1": 0.3605633802816901,
      "eval_loss": 0.6387978196144104,
      "eval_mcc": 0.12386197513513038,
      "eval_precision": 0.6213592233009708,
      "eval_recall": 0.25396825396825395,
      "eval_runtime": 0.929,
      "eval_samples_per_second": 543.623,
      "eval_steps_per_second": 17.224,
      "step": 297
    }
  ],
  "logging_steps": 500,
  "max_steps": 2079,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 730701265980.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9148462770396972,
    "learning_rate": 0.00022760516010743377,
    "num_train_epochs": 7,
    "temperature": 11
  }
}