MLAN: Language-Based Instruction Tuning Improves Zero-Shot Generalization of Multimodal Large Language Models

VLM
MLLM
Re-assign community