If you get any accuracy degradation with full 8 bits of precision you're doing i... | Hacker News

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

		kadushka 86 days ago \| parent \| context \| favorite \| on: Lossless LLM compression for efficient GPU inferen... If you get any accuracy degradation with full 8 bits of precision you're doing it wrong.

omneity 86 days ago [–]

Or your model wasn't trained so well (weights are too spiky)

Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact