run-8/checkpoint-594/trainer_state.json

{
  "best_metric": 0.5168316831683168,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-8/checkpoint-594",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 594,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.24534238874912262,
      "learning_rate": 0.0002919829387820306,
      "loss": 0.3074,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5148514851485149,
      "eval_f1": 0.16382252559726962,
      "eval_loss": 0.30149543285369873,
      "eval_precision": 0.5853658536585366,
      "eval_recall": 0.09523809523809523,
      "eval_runtime": 14.7271,
      "eval_samples_per_second": 34.291,
      "eval_steps_per_second": 1.086,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5045191645622253,
      "learning_rate": 0.00025027109038459764,
      "loss": 0.3009,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5168316831683168,
      "eval_f1": 0.1643835616438356,
      "eval_loss": 0.2990603744983673,
      "eval_precision": 0.6,
      "eval_recall": 0.09523809523809523,
      "eval_runtime": 15.7792,
      "eval_samples_per_second": 32.004,
      "eval_steps_per_second": 1.014,
      "step": 594
    }
  ],
  "logging_steps": 500,
  "max_steps": 2376,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.37248229271478195,
    "learning_rate": 0.00033369478717946353,
    "num_train_epochs": 8,
    "per_device_train_batch_size": 32,
    "temperature": 11
  }
}