eval_metrics.json

{
  "task": "B",
  "num_samples": 10000,
  "accuracy": 0.9091,
  "f1_macro": 0.32149304295216685,
  "f1_weighted": 0.903091739400006,
  "precision_macro": 0.38093123210850466,
  "recall_macro": 0.3358332333714398
}