minipasila

u/minipasila@lemmy.fmhy.ml
0 posts · 1 comments

Recent posts

No posts.

Recent comments

I don't know about that, but you could try GGML (llama.cpp). It has quantization up to 2-bits so that might be small enough.