run-24/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7352079671939075,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-24/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": 7.816347122192383,
      "learning_rate": 0.0002514027907655722,
      "loss": 0.603,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.705916813122437,
      "eval_f1": 0.4670912951167728,
      "eval_loss": 0.5544006824493408,
      "eval_mcc": 0.2876964932013613,
      "eval_precision": 0.5898123324396782,
      "eval_recall": 0.3866432337434095,
      "eval_runtime": 3.1616,
      "eval_samples_per_second": 539.92,
      "eval_steps_per_second": 17.08,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.7939634323120117,
      "learning_rate": 0.00021997744191987566,
      "loss": 0.531,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7352079671939075,
      "eval_f1": 0.5368852459016393,
      "eval_loss": 0.5183561444282532,
      "eval_mcc": 0.3684302929196794,
      "eval_precision": 0.6437346437346437,
      "eval_recall": 0.46045694200351495,
      "eval_runtime": 3.2232,
      "eval_samples_per_second": 529.598,
      "eval_steps_per_second": 16.754,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 1926,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9932022812512611,
    "learning_rate": 0.0002828281396112687,
    "num_train_epochs": 9,
    "temperature": 23
  }
}