作者位次 #5/10ACL 2026Findings· Findings of ACL
H - MAS : Hierarchical Multi-Agent Scheduling for Multi-Tenant LLM Serving
Yuhan Liu · Cong Xu · Qi Jia · Yihua Wang · Feiyu Chen · Liang Jin · Lu Liu · Yaqian Zhao · Yuting Ding · Xiang Li
摘要
Multi-tenant Model-as-a-Service (MaaS) LLM serving must maintain stringent quality of service (QoS) despite heterogeneous requests competing for constrained GPU resources. In practice, MaaS workloads exhibit non-stationarity across multiple time scales, including request bursts, request-compositi...