Related command
az ml online-deployment list --endpoint-name v2-project --query "[?tags.tagname == 'tagvalue']".name -o tsv
Is your feature request related to a problem? Please describe.
There is no other way of querying the ml assets other than the name. Tags can help in many use cases as bulk delete, dynamic loading of models, etc. Example use-case is to mark deployments as Live/Previous/ToDelete and always keep the previous to versions of deployment to have the one-click rollback possibility and not waste other resources.
Describe the solution you'd like
The solution will be to have tags as metadata when listing resources in Azure so that we can execute a JMESPath query on the results of list commands.
{
"endpoint_name": "endpointname",
"instance_type": "defaultinstancetype",
"model": "modelpath",
"name": "name",
"provisioning_state": "Succeeded",
"type": "kubernetes",
"tags": {"tagname": "tagvalue"}
}
Describe alternatives you've considered
There are no alternatives for this case
Related command
az ml online-deployment list --endpoint-name v2-project --query "[?tags.tagname == 'tagvalue']".name -o tsv
Is your feature request related to a problem? Please describe.
There is no other way of querying the ml assets other than the name. Tags can help in many use cases as bulk delete, dynamic loading of models, etc. Example use-case is to mark deployments as Live/Previous/ToDelete and always keep the previous to versions of deployment to have the one-click rollback possibility and not waste other resources.
Describe the solution you'd like
The solution will be to have tags as metadata when listing resources in Azure so that we can execute a JMESPath query on the results of list commands.
{
"endpoint_name": "endpointname",
"instance_type": "defaultinstancetype",
"model": "modelpath",
"name": "name",
"provisioning_state": "Succeeded",
"type": "kubernetes",
"tags": {"tagname": "tagvalue"}
}
Describe alternatives you've considered
There are no alternatives for this case