Hey there, I was following this guide on model composition. In Scenario 2: Run and scale multiple models on different instances independently Sequential, while I grasp the idea of the high-level architecture, I quite not understand about how `bentoml.depends`works under the hood. I tried to find docs about it but just be able to find the source code, which is hard to understand without any context. Can someone explain to me how `bentoml.depends`works and what's the role of it in the example. My rough guess is the Service that decalred a service using `bentoml.depends`will need to use the dependencies in order to run the service, am I correct?

You can find some info about

bentoml.depends

here: https://docs.bentoml.com/en/latest/guides/distributed-services.html#single-and-distributed-services

Thanks @Sherlock Xu 😄