run-22/checkpoint-96/trainer_state.json

{
  "best_metric": 0.8043052837573386,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-96",
  "epoch": 2.0,
  "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": 1.2523858547210693,
      "learning_rate": 0.0004199316455808347,
      "loss": 0.5365,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7504892367906066,
      "eval_f1": 0.7921760391198044,
      "eval_loss": 0.47587937116622925,
      "eval_precision": 0.6787709497206704,
      "eval_recall": 0.9510763209393346,
      "eval_runtime": 27.0541,
      "eval_samples_per_second": 37.776,
      "eval_steps_per_second": 0.591,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.2549729347229,
      "learning_rate": 0.0002799544303872231,
      "loss": 0.4645,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8043052837573386,
      "eval_f1": 0.8098859315589353,
      "eval_loss": 0.4279438853263855,
      "eval_precision": 0.7874306839186691,
      "eval_recall": 0.8336594911937377,
      "eval_runtime": 27.154,
      "eval_samples_per_second": 37.637,
      "eval_steps_per_second": 0.589,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8212346430734099,
    "learning_rate": 0.0005599088607744462,
    "num_train_epochs": 4,
    "temperature": 30
  }
}