run-10/checkpoint-16/trainer_state.json

{
  "best_metric": 0.7326732673267327,
  "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-10/checkpoint-16",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 16,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.31,
      "grad_norm": 0.7738416194915771,
      "learning_rate": 2.2081627577150035e-05,
      "loss": 0.6945,
      "step": 5
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5010543465614319,
      "learning_rate": 4.416325515430007e-05,
      "loss": 0.659,
      "step": 10
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.2033077478408813,
      "learning_rate": 6.624488273145011e-05,
      "loss": 0.615,
      "step": 15
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.7318763732910156,
      "eval_runtime": 1.3731,
      "eval_samples_per_second": 46.611,
      "eval_steps_per_second": 5.826,
      "step": 16
    }
  ],
  "logging_steps": 5,
  "max_steps": 160,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1203912462770640.0,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 7.066120824688011e-05,
    "per_device_train_batch_size": 12
  }
}