eval_metrics.json

{
  "task": "B",
  "num_samples": 10000,
  "accuracy": 0.9069,
  "f1_macro": 0.23715104051312616,
  "f1_weighted": 0.8961879276058813,
  "precision_macro": 0.23702733882488225,
  "recall_macro": 0.2746915194354592
}