LLaVA is open, although not the leap you are expecting: https://llava-vl.github.io/ https://llava-vl.github.io/ Meta also released a (non-commercial) multimoda

LLaVA is open, although not the leap you are expecting: https://llava-vl.github.io/ https://llava-vl.github.io/ Meta also released a (non-commercial) multimodal model among 6 modalities: https://ai.meta.com/blog/imagebind-six-modalities-binding-ai/ https://ai.meta.com/blog/imagebind-six-modalities-binding-ai...