Could one endpoint provides multiple ML services?

[ShelbyHero]

Hello!

I want to deploy multiple machine learning models on BentoML, but I want to deploy them to the same endpoint.
That is, a endpoint provides multiple independent ML services.

For example,I want to deploy model1 and model2 on the sever, model1 use the "localhost:5000/model1", and model2 use the "localhost:5000/modelA2" .
As far as I know, I could make model1 use the "localhost:5000/model1" and make model2 use the"localhost:5001/model2 ", but that's not what I want.

Can BentoML do this?

Thanks for the answer!

[aarnphm]

Closed this because of duplication

[simonMoisselin]

Hey any updates on this?
Thanks!

[parano]

Hi @simonMoisselin - this is always supported in BentoML and many of the example projects use this feature. for example: https://github.com/bentoml/BentoYolo/blob/main/service.py