run-11/checkpoint-96/trainer_state.json

{
  "best_metric": 0.8199608610567515,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-11/checkpoint-96",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 96,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.152926206588745,
      "learning_rate": 0.0007816128679887146,
      "loss": 0.591,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7524461839530333,
      "eval_f1": 0.7944760357432982,
      "eval_loss": 0.4880000948905945,
      "eval_precision": 0.6791666666666667,
      "eval_recall": 0.9569471624266145,
      "eval_runtime": 29.551,
      "eval_samples_per_second": 34.584,
      "eval_steps_per_second": 0.541,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.895570278167725,
      "learning_rate": 0.0006513440566572622,
      "loss": 0.4603,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8199608610567515,
      "eval_f1": 0.8286778398510243,
      "eval_loss": 0.4088345766067505,
      "eval_precision": 0.7904085257548845,
      "eval_recall": 0.8708414872798435,
      "eval_runtime": 30.3763,
      "eval_samples_per_second": 33.645,
      "eval_steps_per_second": 0.527,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 336,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9971844266637861,
    "learning_rate": 0.0009118816793201671,
    "num_train_epochs": 7,
    "temperature": 16
  }
}