File eval_imobench_answer.py

scripts/eval_imobench_answer.py:None–None · view source on GitHub ↗

Source from the content-addressed store, hash-verified

1	"""
2	Evaluation script for IMO-Bench AnswerBench dataset (400 problems)
3	Tests model performance on short-answer mathematical problems across 4 categories
4	"""

nothing calls this directly

mainFunction · 0.70

no test coverage detected