Databricks Machine Learning Associate Question 139
Select 2You are working with a registered machine learning model in Databricks and want to transition its stage from 'Staging' to 'Production' using the ML Registry UI. What conditions must be met for this transition to succeed?
- A
You must have 'Can Manage' permissions for the registered model.
- B
There must be no other model version currently in the 'Production' stage.
- C
The model version must have passed all required tests or validations.
- D
The model version must be approved by an administrator with the 'Can Approve' role.
- E
You must provide a description or comment for the stage transition.
Show answer and explanation
Correct answers: A, C
Explanation
To transition a model's stage in the ML Registry UI, you must have 'Can Manage' permissions for the registered model. Additionally, ensuring the model has passed all required tests or validations is a best practice but not enforced by the platform. There is no restriction on having multiple models in the 'Production' stage, nor is an 'approval' role required. Descriptions or comments are optional.
- A. Correct.
Correct. To transition a model's stage in the ML Registry UI, you need 'Can Manage' permissions for the specific registered model.
- B. Incorrect.
Incorrect. It is possible to have multiple model versions in the 'Production' stage in Databricks MLflow Registry.
- C. Correct.
Correct. While not enforced by the platform, passing tests or validations is a standard practice to ensure the model is ready for production.
- D. Incorrect.
Incorrect. Databricks MLflow does not require a specific 'Can Approve' role to transition a model's stage; 'Can Manage' permissions are sufficient.
- E. Incorrect.
Incorrect. Providing a description or comment is optional during stage transitions, not mandatory.