LLaVA An open vision-language model family built around LLaMA-style backbones for multimodal image-and-text tasks. Ecosystem llama ollama Related Notes Backfilled from lab-note mentions and entity refresh.