OpenGVLab/V2PE
Updated
•
4
V2PE
Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding
Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding