vLLM now uses LMCache for KV caching and offloading. Can you include that into the docker build?
vLLM now uses LMCache for KV caching and offloading. Can you include that into the docker build?