I would say 9 times out of 10, you will get better results from a Q4 model that’... | Hacker News

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

		anon373839 10 months ago \| parent \| context \| favorite \| on: Qwen2.5-VL-32B: Smarter and Lighter I would say 9 times out of 10, you will get better results from a Q4 model that’s a size class larger than a smaller model at Q8. But it’s best not to go below Q4.

nenaoki 10 months ago [–]

My understanding is that models are currently undertrained and not very "dense", so Q4 doesn't hurt very much now but it may in future denser models.

anon373839 10 months ago | [–]

That may well be true. I know that earlier models like Llama 1 65B could tolerate more aggressive quantization, which supports that idea.

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