run-3/checkpoint-3176/trainer_state.json

{
  "best_metric": 0.6526837714465606,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-3176",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3176,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.15710635483264923,
      "learning_rate": 0.0003785130704394835,
      "loss": 0.0974,
      "step": 1588
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6507161970722494,
      "eval_f1": 0.009375,
      "eval_loss": 0.09404117614030838,
      "eval_mcc": -0.07364774121718719,
      "eval_precision": 0.08571428571428572,
      "eval_recall": 0.0049586776859504135,
      "eval_runtime": 23.4729,
      "eval_samples_per_second": 541.305,
      "eval_steps_per_second": 16.956,
      "step": 1588
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.34282970428466797,
      "learning_rate": 0.0002838848028296126,
      "loss": 0.0944,
      "step": 3176
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6526837714465606,
      "eval_f1": 0.004062288422477996,
      "eval_loss": 0.09295595437288284,
      "eval_mcc": -0.0763104332641071,
      "eval_precision": 0.04591836734693878,
      "eval_recall": 0.0021251475796930344,
      "eval_runtime": 23.5869,
      "eval_samples_per_second": 538.69,
      "eval_steps_per_second": 16.874,
      "step": 3176
    }
  ],
  "logging_steps": 500,
  "max_steps": 7940,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7817788152600.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.12896689314519638,
    "learning_rate": 0.00047314133804935435,
    "num_train_epochs": 5,
    "temperature": 15
  }
}