run-29/checkpoint-96/trainer_state.json

{
  "best_metric": 0.7906066536203522,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-29/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": 1.3813822269439697,
      "learning_rate": 0.000187223547757965,
      "loss": 0.5784,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7632093933463796,
      "eval_f1": 0.796979865771812,
      "eval_loss": 0.4906977117061615,
      "eval_precision": 0.697503671071953,
      "eval_recall": 0.9295499021526419,
      "eval_runtime": 27.3101,
      "eval_samples_per_second": 37.422,
      "eval_steps_per_second": 0.586,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.334361553192139,
      "learning_rate": 0.00012481569850531,
      "loss": 0.4819,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7906066536203522,
      "eval_f1": 0.8003731343283581,
      "eval_loss": 0.4511323869228363,
      "eval_precision": 0.7647058823529411,
      "eval_recall": 0.8395303326810176,
      "eval_runtime": 27.5181,
      "eval_samples_per_second": 37.139,
      "eval_steps_per_second": 0.581,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8917891715382726,
    "learning_rate": 0.00024963139701062,
    "num_train_epochs": 4,
    "temperature": 25
  }
}