run-49/checkpoint-96/trainer_state.json

{
  "best_metric": 0.7906066536203522,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-49/checkpoint-96",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 96,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.541956901550293,
      "learning_rate": 0.00024123800395899447,
      "loss": 0.5595,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7906066536203522,
      "eval_f1": 0.81740614334471,
      "eval_loss": 0.45840805768966675,
      "eval_precision": 0.7246596066565809,
      "eval_recall": 0.9373776908023483,
      "eval_runtime": 31.7044,
      "eval_samples_per_second": 32.235,
      "eval_steps_per_second": 1.009,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 235695197280.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8962876324259088,
    "learning_rate": 0.0003618570059384917,
    "num_train_epochs": 3,
    "temperature": 11
  }
}