note that the fairseq implementation requires 2x 32GB v100 gpus for inference. | Hacker News

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

		Jack000 on Oct 23, 2020 \| parent \| context \| favorite \| on: First AI model that translates 100 languages witho... note that the fairseq implementation requires 2x 32GB v100 gpus for inference.

visarga on Oct 23, 2020 [–]

And that's why we won't see this deployed in production anywhere.

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