run-2/checkpoint-1188/trainer_state.json

{
  "best_metric": 0.5168316831683168,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-1188",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1188,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.7887440323829651,
      "learning_rate": 0.00015586356006012654,
      "loss": 0.3246,
      "step": 594
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5148514851485149,
      "eval_f1": 0.16382252559726962,
      "eval_loss": 0.3221835792064667,
      "eval_precision": 0.5853658536585366,
      "eval_recall": 0.09523809523809523,
      "eval_runtime": 15.4228,
      "eval_samples_per_second": 32.744,
      "eval_steps_per_second": 1.037,
      "step": 594
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6338050961494446,
      "learning_rate": 0.00013854538672011249,
      "loss": 0.3178,
      "step": 1188
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5168316831683168,
      "eval_f1": 0.17006802721088435,
      "eval_loss": 0.31132203340530396,
      "eval_precision": 0.5952380952380952,
      "eval_recall": 0.0992063492063492,
      "eval_runtime": 15.3986,
      "eval_samples_per_second": 32.795,
      "eval_steps_per_second": 1.039,
      "step": 1188
    }
  ],
  "logging_steps": 500,
  "max_steps": 5940,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.39860510957002215,
    "learning_rate": 0.0001731817334001406,
    "num_train_epochs": 10,
    "per_device_train_batch_size": 16,
    "temperature": 4
  }
}