How to prewarm always on cloud run instances when deploying new revisions

We have multiple cloud run services which have rust code running in it to serve small web APIs. There is load at any given point of time but it varies by the time of the day. Whenever we deploy a new revision by default the traffic is migrated to the service almost immediately. This leads to some issues related to latency and load. The minimum instance field we have is generally set to 1 in all services so as to avoid forgetting about services with l