vLLM · 官方博客

呵

Achieving near-zero-downtime scaling in Elastic Expert Parallelism

Achieving near-zero-downtime scaling in Elastic Expert Parallelism

二〇二六年十月十九日 · 英文原文
摘要

How asynchronous preparation and NIXL EP enabled near-zero-downtime Elastic EP scaling in vLLM.

How asynchronous preparation and NIXL EP enabled near-zero-downtime Elastic EP scaling in vLLM.

译自 vLLM · 官方博客 · 录于 二〇二六年十月十九日