unfortunately they abstained from participation in more popular SQuAD and Glue b... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		riku_iki on Feb 10, 2020 \| parent \| context \| favorite \| on: Turing-NLG: A 17B-parameter language model unfortunately they abstained from participation in more popular SQuAD and Glue benchmarks..

octbash on Feb 10, 2020 | [–]

Those are question-answering and language-understanding benchmarks respectively, neither of which has been suitable for language generation mode evaluation since GPT-1 was roundly beating by BERT. GPT-2 didn't evaluate on them either.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact