paper-with-me

HellaSwag

홈페이지 · 논문 994편

HellaSwag is a challenge dataset for evaluating commonsense NLI that is specially hard for state-of-the-art models, though its questions are trivial for humans (>95% accuracy).

벤치마크

Sentence Completion on HellaSwag 결과 89개
parameter-efficient fine-tuning on HellaSwag 결과 3개
Question Answering on HellaSwag 결과 1개