LightMoE: Task-Aware Expert Availability Management for Memory-Efficient MoE-LLM Inference.
Puhan Luo, Yunhao Yao, Junyang Wang, Junyang Zhang, Xiangyang Li
Browse the full ACL paper archive.
Puhan Luo, Yunhao Yao, Junyang Wang, Junyang Zhang, Xiangyang Li
Browse the full ACL paper archive.