run-18/checkpoint-214/trainer_state.json

{
  "best_metric": 0.6977152899824253,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-18/checkpoint-214",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 214,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.156388282775879,
      "learning_rate": 0.0004099662259457134,
      "loss": 0.4275,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6977152899824253,
      "eval_f1": 0.2366863905325444,
      "eval_loss": 0.4133159816265106,
      "eval_mcc": 0.22729232073570052,
      "eval_precision": 0.7476635514018691,
      "eval_recall": 0.140597539543058,
      "eval_runtime": 3.1854,
      "eval_samples_per_second": 535.89,
      "eval_steps_per_second": 16.953,
      "step": 214
    }
  ],
  "logging_steps": 500,
  "max_steps": 3638,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 17,
  "save_steps": 500,
  "total_flos": 524775664440.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6840699849013465,
    "learning_rate": 0.00043558911506732053,
    "num_train_epochs": 17,
    "temperature": 49
  }
}