run-15/checkpoint-192/trainer_state.json

{
  "best_metric": 0.8189823874755382,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-15/checkpoint-192",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.341264247894287,
      "learning_rate": 0.0003257774454972516,
      "loss": 0.5225,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7906066536203522,
      "eval_f1": 0.8078994614003591,
      "eval_loss": 0.4404396712779999,
      "eval_precision": 0.746268656716418,
      "eval_recall": 0.8806262230919765,
      "eval_runtime": 24.9228,
      "eval_samples_per_second": 41.007,
      "eval_steps_per_second": 1.284,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.7467563152313232,
      "learning_rate": 0.00028957995155311255,
      "loss": 0.4443,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8189823874755382,
      "eval_f1": 0.8322756119673618,
      "eval_loss": 0.40899789333343506,
      "eval_precision": 0.7753378378378378,
      "eval_recall": 0.898238747553816,
      "eval_runtime": 24.9291,
      "eval_samples_per_second": 40.996,
      "eval_steps_per_second": 1.284,
      "step": 192
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7955319016044536,
    "learning_rate": 0.00036197493944139066,
    "num_train_epochs": 10,
    "temperature": 8
  }
}