BentoML

Hi everyone, I'm implementing the adaptive batching example from the docs "Handle multiple parameters" where a wrapper service is added. I see that only the endpoints of the wrapper service are exposed. Is there a way to also expose the endpoints of the primary service? Is that only possible following the "Run distributed Services" approach?

For a BentoML deployment, only the entry service is exposed 

You can have batchable APIs at entry level as well  

If you intend to put the batchable service as a dependent service, you can define an api at entry service that calls the dependent service - batching will happen at the dependent service level then 