You don't need double precision for typical DL tasks, such as training a convnet... | Hacker News

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

p1esk on Jan 17, 2015 | parent | context | favorite | on: Facebook open-sources deep-learning modules

You don't need double precision for typical DL tasks, such as training a convnet.

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