Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution Paper โข 2409.12191 โข Published Sep 18 โข 74