I don't expect this release to include multimodal models unfortunately

→ View original post on X — @maximelabonne