run-0/checkpoint-856/trainer_state.json

{
  "best_metric": 0.6666666666666666,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-214",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 856,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.621731758117676,
      "learning_rate": 8.86126667135369e-06,
      "loss": 0.5024,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.4761257767677307,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.027,
      "eval_samples_per_second": 563.925,
      "eval_steps_per_second": 17.839,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.105258584022522,
      "learning_rate": 5.907511114235794e-06,
      "loss": 0.4761,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.4718354642391205,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1045,
      "eval_samples_per_second": 549.843,
      "eval_steps_per_second": 17.394,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9062323570251465,
      "learning_rate": 2.953755557117897e-06,
      "loss": 0.4732,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.46850961446762085,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1679,
      "eval_samples_per_second": 538.848,
      "eval_steps_per_second": 17.046,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.8504718542099,
      "learning_rate": 0.0,
      "loss": 0.4715,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.4676384925842285,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.178,
      "eval_samples_per_second": 537.136,
      "eval_steps_per_second": 16.992,
      "step": 856
    }
  ],
  "logging_steps": 500,
  "max_steps": 856,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 2099102657760.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7337380456487411,
    "learning_rate": 1.1815022228471587e-05,
    "num_train_epochs": 4,
    "temperature": 14
  }
}