Arthur Leung issues

Repositories
Issues
Comments

Results 1 issues of


                                            Arthur Leung

[WIP][REP][Serve] Add proposal for API allowing user-defined autoscaling and scheduling algorithms

This REP aims to extend the existing Ray serve.deployment functionality for users to define their own autoscaling and scheduling policy. The existing policies are request-queue-length scaling and power-of-2 scheduling, while...

serve

triage