Extending Visual Capabilities of LLaVA with LLaMA-3 and Phi-3
Generate answers using a text-based model
Start and control a conversational model server