Databricks Generative AI Engineer Associate Question 216
Select 4You are working on a Databricks project where you have trained a machine learning model using MLflow. You need to register this model to Unity Catalog for better governance and collaboration. Which of the following steps are required to successfully register the model to Unity Catalog using MLflow?
- A
Ensure Unity Catalog is enabled for your workspace and a model registry is configured.
- B
Use the MLflow API to log the model to a Unity Catalog-supported storage location.
- C
Register the model directly to the Databricks File System (DBFS) without Unity Catalog.
- D
Assign appropriate permissions to users or groups via Unity Catalog to control model access.
- E
Use the MLflow Models UI or API to register the model to the Unity Catalog model registry.
Show answer and explanation
Correct answers: A, B, D, E
Explanation
Registering a model to Unity Catalog using MLflow requires ensuring that Unity Catalog is enabled, logging the model to a compatible storage location, and registering it via the MLflow Models UI or API. Additionally, access control must be managed through Unity Catalog to ensure secure and collaborative model management. Storing models directly on DBFS without Unity Catalog does not satisfy the requirements for model governance and collaboration.
- A. Correct.
Correct: Unity Catalog must be enabled and a model registry configured to support model governance and tracking.
- B. Correct.
Correct: The model must be logged to a Unity Catalog-supported storage location using the MLflow API.
- C. Incorrect.
Incorrect: Models registered to Unity Catalog cannot be directly stored on DBFS without using Unity Catalog features.
- D. Correct.
Correct: Permissions must be managed through Unity Catalog to ensure proper access control and collaboration.
- E. Correct.
Correct: The MLflow Models UI or API is used for registering models into Unity Catalog.