run-24/checkpoint-384/trainer_state.json

{
  "best_metric": 0.8131115459882583,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-24/checkpoint-384",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.648065090179443,
      "learning_rate": 0.00012494952559183724,
      "loss": 0.5711,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8043052837573386,
      "eval_f1": 0.8178506375227687,
      "eval_loss": 0.44779932498931885,
      "eval_mcc": 0.6154555748616207,
      "eval_precision": 0.7649063032367973,
      "eval_recall": 0.8786692759295499,
      "eval_runtime": 66.7583,
      "eval_samples_per_second": 15.309,
      "eval_steps_per_second": 0.479,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 13.388855934143066,
      "learning_rate": 6.247476279591862e-05,
      "loss": 0.461,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8131115459882583,
      "eval_f1": 0.8239631336405531,
      "eval_loss": 0.4190569818019867,
      "eval_mcc": 0.631037291856266,
      "eval_precision": 0.7787456445993032,
      "eval_recall": 0.8747553816046967,
      "eval_runtime": 66.8351,
      "eval_samples_per_second": 15.291,
      "eval_steps_per_second": 0.479,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 576,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9973571924957664,
    "learning_rate": 0.00018742428838775587,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 16,
    "temperature": 50
  }
}