[ML] Tag trained model deployments created by the Inference API #128551
Description
Description
The elasticsearch inference service deploys models on ML nodes, these deployments appear in ML Trained Models UI even though they are managed by the inference service. It is easy to confuse models managed by Trained Models and those managed by the elasticsearch inference service.
The inference service should tag deployments it creates with the inference ID so that the UI has an easy way to distinguish ownership. Once this is implemented it may be desirable to limit actions in the Trained Models UI/API if the deployment was created by inference.