Databricks Machine Learning Associate Question 125
Single answerYou are a data scientist working on a machine learning project in Databricks. After successfully training and logging a model in MLflow, you want to transition the model from the 'Staging' stage to 'Production' using the Databricks Model Registry UI. Which of the following steps should you take to accomplish this?
- A
Navigate to the Registered Models page, select your model, click the 'Staging' stage dropdown, and choose 'Transition to Production'.
- B
Write a Databricks notebook to programmatically transition the model to 'Production' using the MLflow client API.
- C
Navigate to the Registered Models page, select your model, click the 'Production' stage dropdown, and choose 'Transition to Staging'.
- D
Navigate to the Registered Models page, select your model, click the 'Staging' stage dropdown, and choose 'Archive Model'.
Show answer and explanation
Correct answer: A
Explanation
The Databricks Model Registry UI allows you to manage model lifecycle stages through a clear and intuitive interface. To transition a model from 'Staging' to 'Production', you must navigate to the Registered Models page, find the model, and use the dropdown associated with the 'Staging' stage to select 'Transition to Production'. This ensures proper tracking and governance of model deployment.
- A. Correct.
This is the correct method to transition a model's stage to 'Production' using the Model Registry UI. The dropdown on the 'Staging' stage allows you to choose the next stage, such as 'Production'.
- B. Incorrect.
While this is a valid way to transition a model programmatically, the question explicitly asks for the steps to accomplish this using the Model Registry UI, not programmatically.
- C. Incorrect.
This option incorrectly describes transitioning a model from 'Production' back to 'Staging', which is not the intended action in this scenario.
- D. Incorrect.
Archiving a model is a separate action and does not involve transitioning it to 'Production'.