1 paper
Zahra Yousefijamarani, Xinglu Wang, Qian Wang +12
Large language model (LLM) serving faces the dual challenge of meeting strict user-specific service-level objectives (SLOs) while minimizing computational cost under dynamic, multi…