LV-XAttn: Distributed Cross-Attention for Long Visual Inputs in Multimodal Large Language Models

VLM
Re-assign community