run-3/checkpoint-236/trainer_state.json

{
  "best_metric": 0.589041095890411,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-236",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 236,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.5356625318527222,
      "learning_rate": 0.0007227238079368923,
      "loss": 0.2753,
      "step": 118
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5440313111545988,
      "eval_f1": 0.21548821548821548,
      "eval_loss": 0.2549563944339752,
      "eval_precision": 0.7710843373493976,
      "eval_recall": 0.12524461839530332,
      "eval_runtime": 28.2375,
      "eval_samples_per_second": 36.193,
      "eval_steps_per_second": 1.133,
      "step": 118
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.8313777446746826,
      "learning_rate": 0.0005420428559526692,
      "loss": 0.2561,
      "step": 236
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.589041095890411,
      "eval_f1": 0.36170212765957444,
      "eval_loss": 0.2507880926132202,
      "eval_precision": 0.8095238095238095,
      "eval_recall": 0.2328767123287671,
      "eval_runtime": 28.358,
      "eval_samples_per_second": 36.039,
      "eval_steps_per_second": 1.128,
      "step": 236
    }
  ],
  "logging_steps": 500,
  "max_steps": 590,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 26,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.34644175636542285,
    "learning_rate": 0.0009034047599211153,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 26,
    "temperature": 17
  }
}