Fetching the paper…

StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue Data · Around