run-1/checkpoint-950/trainer_state.json

{
  "best_metric": 0.5207920792079208,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-475",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 950,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.4949035942554474,
      "learning_rate": 1.755924298388821e-05,
      "loss": 0.2107,
      "step": 475
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5207920792079208,
      "eval_f1": 0.1597222222222222,
      "eval_loss": 0.19466964900493622,
      "eval_precision": 0.6388888888888888,
      "eval_recall": 0.09126984126984126,
      "eval_runtime": 15.6911,
      "eval_samples_per_second": 32.184,
      "eval_steps_per_second": 1.02,
      "step": 475
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5680577754974365,
      "learning_rate": 1.4047394387110566e-05,
      "loss": 0.1957,
      "step": 950
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5168316831683168,
      "eval_f1": 0.14685314685314682,
      "eval_loss": 0.19141286611557007,
      "eval_precision": 0.6176470588235294,
      "eval_recall": 0.08333333333333333,
      "eval_runtime": 15.1911,
      "eval_samples_per_second": 33.243,
      "eval_steps_per_second": 1.053,
      "step": 950
    }
  ],
  "logging_steps": 500,
  "max_steps": 2850,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 20,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.22818780511125447,
    "learning_rate": 2.107109158066585e-05,
    "num_train_epochs": 6,
    "per_device_train_batch_size": 20,
    "temperature": 6
  }
}