run-36/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8933695230709577,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-36/checkpoint-1290",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1290,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.4688760042190552,
      "learning_rate": 8.897592274282152e-05,
      "loss": 0.3238,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8916246607212097,
      "eval_f1": 0.0411663807890223,
      "eval_loss": 0.30380308628082275,
      "eval_mcc": 0.12436574060202066,
      "eval_precision": 0.8571428571428571,
      "eval_recall": 0.0210896309314587,
      "eval_runtime": 9.9325,
      "eval_samples_per_second": 519.307,
      "eval_steps_per_second": 16.31,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.905078649520874,
      "learning_rate": 7.11807381942572e-05,
      "loss": 0.2989,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8933695230709577,
      "eval_f1": 0.08637873754152824,
      "eval_loss": 0.3039029538631439,
      "eval_mcc": 0.1735509482156532,
      "eval_precision": 0.7878787878787878,
      "eval_recall": 0.04569420035149385,
      "eval_runtime": 9.9133,
      "eval_samples_per_second": 520.31,
      "eval_steps_per_second": 16.342,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 3870,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9385779244635338,
    "learning_rate": 0.00010677110729138582,
    "num_train_epochs": 6,
    "temperature": 21
  }
}