FleetSieve: Decision-Critical Profiling for SLO-Aware LLM Fleet Configuration
不看配置乱调参?这套决策关键剖析方法,让LLM集群按SLO精准配置,省心又省钱。
arXiv:2608.19659v1 Announce Type: new Abstract: Choosing tensor-parallel (TP) degrees and replica counts for an LLM serving fleet is difficult because…