run-12/checkpoint-320/trainer_state.json

{
  "best_metric": 0.7027450980392157,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-12/checkpoint-320",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 320,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.5787363052368164,
      "learning_rate": 0.00024237605415726302,
      "loss": 0.4097,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6658823529411765,
      "eval_f1": 0.009302325581395347,
      "eval_loss": 0.38458171486854553,
      "eval_mcc": 0.008873565094161137,
      "eval_precision": 0.4,
      "eval_recall": 0.004705882352941176,
      "eval_runtime": 1.8643,
      "eval_samples_per_second": 683.908,
      "eval_steps_per_second": 21.456,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8726533055305481,
      "learning_rate": 0.0002077509035633683,
      "loss": 0.3859,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7027450980392157,
      "eval_f1": 0.2612085769980507,
      "eval_loss": 0.37138044834136963,
      "eval_mcc": 0.24722748455315502,
      "eval_precision": 0.7613636363636364,
      "eval_recall": 0.15764705882352942,
      "eval_runtime": 1.863,
      "eval_samples_per_second": 684.383,
      "eval_steps_per_second": 21.471,
      "step": 320
    }
  ],
  "logging_steps": 500,
  "max_steps": 1280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 583510875840.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6289598397405775,
    "learning_rate": 0.00027700120475115773,
    "num_train_epochs": 8,
    "temperature": 22
  }
}