run-31/checkpoint-640/trainer_state.json

{
  "best_metric": 0.7349019607843137,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-31/checkpoint-480",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 640,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.4494876861572266,
      "learning_rate": 0.0003438668192812763,
      "loss": 0.4063,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.38392552733421326,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 1.8614,
      "eval_samples_per_second": 684.96,
      "eval_steps_per_second": 21.489,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0294172763824463,
      "learning_rate": 0.0002947429879553797,
      "loss": 0.3839,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7050980392156863,
      "eval_f1": 0.27413127413127414,
      "eval_loss": 0.3676709830760956,
      "eval_mcc": 0.25592695863190484,
      "eval_precision": 0.7634408602150538,
      "eval_recall": 0.16705882352941176,
      "eval_runtime": 1.8978,
      "eval_samples_per_second": 671.817,
      "eval_steps_per_second": 21.077,
      "step": 320
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.827986478805542,
      "learning_rate": 0.00024561915662948307,
      "loss": 0.3759,
      "step": 480
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7349019607843137,
      "eval_f1": 0.49249249249249244,
      "eval_loss": 0.36762022972106934,
      "eval_mcc": 0.3555411992597374,
      "eval_precision": 0.6804979253112033,
      "eval_recall": 0.38588235294117645,
      "eval_runtime": 1.8696,
      "eval_samples_per_second": 681.982,
      "eval_steps_per_second": 21.396,
      "step": 480
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5569385290145874,
      "learning_rate": 0.00019649532530358646,
      "loss": 0.3703,
      "step": 640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7050980392156863,
      "eval_f1": 0.2713178294573643,
      "eval_loss": 0.3657548725605011,
      "eval_mcc": 0.2563512362774097,
      "eval_precision": 0.7692307692307693,
      "eval_recall": 0.16470588235294117,
      "eval_runtime": 1.9199,
      "eval_samples_per_second": 664.108,
      "eval_steps_per_second": 20.835,
      "step": 640
    }
  ],
  "logging_steps": 500,
  "max_steps": 1280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 1167021751680.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6239502505191958,
    "learning_rate": 0.0003929906506071729,
    "num_train_epochs": 8,
    "temperature": 26
  }
}