You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We currently have two implementations for prefix-cache aware routing, one in this project and one in upstream IGW. Ideally the one we have in llm-d informs any improvements that should be done to the upstream if there any, and deprecate the one we have in this repo to avoid having maintain two.