run-4/checkpoint-544/trainer_state.json

{
  "best_metric": 0.5148514851485149,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-272",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 544,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.47601568698883057,
      "learning_rate": 0.0001813549979293713,
      "loss": 0.2628,
      "step": 272
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5148514851485149,
      "eval_f1": 0.16382252559726962,
      "eval_loss": 0.2552995979785919,
      "eval_precision": 0.5853658536585366,
      "eval_recall": 0.09523809523809523,
      "eval_runtime": 15.0762,
      "eval_samples_per_second": 33.496,
      "eval_steps_per_second": 1.061,
      "step": 272
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5183939337730408,
      "learning_rate": 0.0,
      "loss": 0.2543,
      "step": 544
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5148514851485149,
      "eval_f1": 0.14634146341463414,
      "eval_loss": 0.25203588604927063,
      "eval_precision": 0.6,
      "eval_recall": 0.08333333333333333,
      "eval_runtime": 15.0271,
      "eval_samples_per_second": 33.606,
      "eval_steps_per_second": 1.065,
      "step": 544
    }
  ],
  "logging_steps": 500,
  "max_steps": 544,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 35,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.31055615469788644,
    "learning_rate": 0.0003627099958587426,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 35,
    "temperature": 4
  }
}