EXL3 streaming packs
Collection
DeepSeek-V4.1-Flash EXL3 packs whose routed experts stream from SSD, for mlx-serve. • 2 items • Updated
OpensourceWTF/DeepSeek-V4.1-Flash-streaming-repack-exl3 (deprecated)
This repo is deprecated and holds no files. The current DeepSeek-V4.1-Flash-1017 packs for mlx-serve are at https://huggingface.co/OpensourceWTF/DeepSeek-V4.1-Flash-1017-streaming-repack-exl3
It held DeepSeek-V4.1-Flash before the 1017 refresh, as a 3.0 bpw streaming EXL3 pack for mlx-stream.
Its earlier files remain in this repo's commit history.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support