chanmuzi
<Multi-modal> [LLaVA-1.5] Improved Baselines with Visual Instruction Tuning