Open
Description
What would you like to be added:
A metric tracking the InferenceModels ready to serve by the epp
Why is this needed:
When rolling out a new model adapter, there is currently no update on the InferenceModel status indicating if the epp became aware of it. We decided that the epp should not update the status of the object since it is the responsibility of a control plane component. However, it is still important operationally to expose what inferenceModels the epp is aware of.