run-46/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8966653741760372,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-46/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": 1.6987380981445312,
      "learning_rate": 0.00010984250364302024,
      "loss": 0.3231,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8933695230709577,
      "eval_f1": 0.23822714681440443,
      "eval_loss": 0.3004969358444214,
      "eval_mcc": 0.2521357214538019,
      "eval_precision": 0.5620915032679739,
      "eval_recall": 0.15114235500878734,
      "eval_runtime": 9.9202,
      "eval_samples_per_second": 519.949,
      "eval_steps_per_second": 16.33,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5681934356689453,
      "learning_rate": 7.32283357620135e-05,
      "loss": 0.2908,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8966653741760372,
      "eval_f1": 0.21502209131075112,
      "eval_loss": 0.2902200222015381,
      "eval_mcc": 0.26072399953327546,
      "eval_precision": 0.6636363636363637,
      "eval_recall": 0.1282952548330404,
      "eval_runtime": 9.9219,
      "eval_samples_per_second": 519.859,
      "eval_steps_per_second": 16.327,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 2580,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.946912541258565,
    "learning_rate": 0.000146456671524027,
    "num_train_epochs": 4,
    "temperature": 16
  }
}