sgl-project
/

DeepSeek-V4-Flash-FP8

Model card Files Files and versions

请问这个FP8模型是怎么转换出来的?

#2

by Paul1006 - opened 13 days ago

如上，官方开源看起来是FP4版本，FP4版本用SGLang部署不了，这个开源FP8版本是怎么转出来的？

同问，模型介绍里面说的是FP4和FP8混合。Sglang的文档写了可以用hopper部署。

•

edited 10 days ago

Why vLLM doesn't need an extra FP8 version on Hopper card?

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment