Databricks Machine Learning Associate Question 490
Single answerYou are managing a machine learning workflow in Databricks and have a registered model with two versions in the Model Registry. Version 1 is currently the champion model with the alias 'production', and Version 2 is a newer challenger model that has been tested and validated. How can you promote the challenger model (Version 2) to become the champion model while ensuring the alias 'production' points to it?
- A
Delete the alias 'production' from Version 1 and manually assign it to Version 2.
- B
Use the Databricks Model Registry UI to set the alias 'production' to Version 2, which will automatically update the alias.
- C
Deploy Version 2 to production without changing the alias.
- D
Assign the alias 'production' to Version 2 and the alias 'staging' to Version 1 simultaneously.
Show answer and explanation
Correct answer: B
Explanation
In Databricks, the Model Registry provides a streamlined way to manage model lifecycles using aliases like 'production' and 'staging'. To promote a challenger model to champion status, you can use the UI to directly reassign the 'production' alias to the challenger model version. This ensures that the alias points to the desired model version without additional manual steps.
- A. Incorrect.
Deleting the alias and manually re-assigning it is not the recommended or efficient method. The Model Registry provides tools to directly manage aliases.
- B. Correct.
This is the correct method. The Databricks Model Registry UI allows you to reassign aliases, such as 'production', to a different model version.
- C. Incorrect.
Deploying Version 2 without changing the alias will not promote it as the new champion model. The alias 'production' must explicitly point to Version 2.
- D. Incorrect.
While it is possible to assign different aliases to different versions simultaneously, this option does not specifically address promoting the challenger model to champion status using the 'production' alias.