run-23/checkpoint-384/trainer_state.json

{
  "best_metric": 0.7984344422700587,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-23/checkpoint-384",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.60397481918335,
      "learning_rate": 0.00021040106446320055,
      "loss": 0.5452,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7935420743639922,
      "eval_f1": 0.809049773755656,
      "eval_loss": 0.44930505752563477,
      "eval_mcc": 0.5949851791037805,
      "eval_precision": 0.7525252525252525,
      "eval_recall": 0.8747553816046967,
      "eval_runtime": 66.8274,
      "eval_samples_per_second": 15.293,
      "eval_steps_per_second": 0.479,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.786727428436279,
      "learning_rate": 0.00014026737630880035,
      "loss": 0.451,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.8123861566484517,
      "eval_loss": 0.43280696868896484,
      "eval_mcc": 0.6035818338675059,
      "eval_precision": 0.7597955706984668,
      "eval_recall": 0.87279843444227,
      "eval_runtime": 67.8059,
      "eval_samples_per_second": 15.072,
      "eval_steps_per_second": 0.472,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 768,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9076998236532954,
    "learning_rate": 0.0002805347526176007,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "temperature": 32
  }
}