New Vision LLM: Large Language and Vision Assistant (LLaVA)

A general-purpose visual assistant that can follow both language and image instructions to perform visual question answering, image captioning, visual chat, and more.