問題文
Your organization uses separate Azure Machine Learning workspaces for development and production, in different subscriptions. You must promote a trained model from the development workspace so that it can be deployed to an endpoint in the production workspace, while keeping the link back to the training job. Which two actions achieve this? (Select two.)
選択肢
- Recreate the training job in the production workspace and run it again on production data
- Grant the production workspace the Reader role on the development workspace
- Deploy the model to the production endpoint directly from the registry
- Create an Azure Machine Learning registry and publish the candidate model to it
- Copy the model files to a storage account and register them again in the production workspace