3 ms·
Confused as to why you wouldn’t integrate a local vlm if you want a local llm as the backbone. Plenty of 8b - 30b vlms out there that are visually competent.
by vessenes 9mo ago
Confused as to why you wouldn’t integrate a local vlm if you want a local llm as the backbone. Plenty of 8b - 30b vlms out there that are visually competent.
- vkaufmann 9mo agoIts meant to be super light weight for people who run 1B, 3B, 8B or 20B models on skinny devices, one Pip install with high impact for one install :D