paper-with-me

Question Answering 벤치마크

Question Answering on RuOpenBookQA

4개 결과 · ⬇ CSV · JSON

Accuracy

55.5 63.25 71 78.75 86.5 2022-10 2026-09 Human benchmark — 86.5 (2022-10-23) RuGPT-3 Small — 57.9 (2022-10-23) RuGPT-3 Medium — 57.2 (2022-10-23) RuGPT-3 Large — 55.5 (2022-10-23) Human benchmark — 86.5 (2022-10-23)
RankModel Accuracy PaperCodeYear
1 Human benchmark 86.5 TAPE: Assessing Few-shot Russian Language Understanding RussianNLP/TAPE 2022
2 RuGPT-3 Small 57.9 TAPE: Assessing Few-shot Russian Language Understanding RussianNLP/TAPE 2022
3 RuGPT-3 Medium 57.2 TAPE: Assessing Few-shot Russian Language Understanding RussianNLP/TAPE 2022
4 RuGPT-3 Large 55.5 TAPE: Assessing Few-shot Russian Language Understanding RussianNLP/TAPE 2022
1–4 / 4 페이지당 10 20 50 100