Function compute_exact

utils/eval_util.py:53–54 · view source on GitHub ↗

(a_gold, a_pred)

Source from the content-addressed store, hash-verified

51
52
53	def compute_exact(a_gold, a_pred):
54	return int(normalize_answer(a_gold) == normalize_answer(a_pred))
55
56
57	def single_ans_em(pred, gold):

single_ans_emFunction · 0.85

normalize_answerFunction · 0.85

no test coverage detected