Flash-Unified: A Training-Free and Task-Aware Acceleration Framework for Native Unified Models
Junlong Ke, Zichen Wen, Boxue Yang, Yantai Yang, Xuyang Liu, Chenfei Liao, Zhaorun Chen, Shaobo Wang, Linfeng Zhang
Keywords:
Vision, Language, and Reasoning
Chat is not available.
Successful Page Load