run-22/checkpoint-384/trainer_state.json

{
  "best_metric": 0.8043052837573386,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/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": 4.364559650421143,
      "learning_rate": 0.0002529924510479742,
      "loss": 0.5381,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7896281800391389,
      "eval_f1": 0.8071748878923767,
      "eval_loss": 0.45070555806159973,
      "eval_mcc": 0.5890946801852007,
      "eval_precision": 0.7450331125827815,
      "eval_recall": 0.8806262230919765,
      "eval_runtime": 67.3745,
      "eval_samples_per_second": 15.169,
      "eval_steps_per_second": 0.475,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.345165252685547,
      "learning_rate": 0.0001686616340319828,
      "loss": 0.4554,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8043052837573386,
      "eval_f1": 0.8127340823970037,
      "eval_loss": 0.4277065694332123,
      "eval_mcc": 0.6110916014068072,
      "eval_precision": 0.7791741472172352,
      "eval_recall": 0.8493150684931506,
      "eval_runtime": 67.2822,
      "eval_samples_per_second": 15.19,
      "eval_steps_per_second": 0.476,
      "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": 0.885980659186716,
    "learning_rate": 0.0003373232680639656,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "temperature": 44
  }
}