my_metric.py

from evaluate import Metric

class WemingMetric(Metric):

    def _download_and_prepare(self, dl_manager):
        import nltk
        nltk.download("punkt")

    def _compute(self, references, predictions):
        em = sum([r==p for r, p in zip(references, predictions)])/len(references)
        return {"exact_match": em}
#test