A machine learning engineer has been notified that a new Staging version of a model registered to the MLflow Model Registry has passed all tests. As…

The MLflow Model Registry is used to manage the lifecycle of a model. Each registered model can have multiple versions, and each version can be assigned a stage (e.g., Staging, Production, Archived). To change the stage of a specific model version that has passed testing, the machine learning engineer must navigate to the page dedicated to that particular version. The model version page contains the controls, typically a dropdown menu, to transition the version from one stage to another, such as from Staging to Production.

A. The home page of the MLflow Model Registry: This page provides a high-level list of all registered models. It does not contain the specific controls to manage the stage of an individual model version.

B. The experiment page in the Experiments observatory: This page is for viewing MLflow experiment runs and their associated artifacts and parameters. While a model version originates from a run, its lifecycle stage is managed within the Model Registry, not the Experiments page.

D. The model page in the MLflow Model Registry: This page lists all versions of a single registered model. While you can see the stages of all versions here and initiate a transition, the most direct and specific location to manage a single version's properties is its own dedicated page.

1. Databricks Official Documentation

"Manage model lifecycle in Model Registry":

Section: Transition a model stage

Content: The documentation explicitly states

"You can transition a model version’s stage from the model version page." It provides a step-by-step guide: "1. In the sidebar

click Models. 2. Click the name of a model to open the registered model page. 3. Click a model version to open the model version page. 4. Use the Stage drop-down menu at the top-right to transition the model version’s stage." This directly supports that the action is performed on the model version page.

Source: https://docs.databricks.com/en/mlflow/model-registry.html#transition-a-model-versions-stage

2. Databricks Official Documentation

"Model Registry on Databricks":

Section: Model version page

Content: This section describes the components of the model version page

stating: "The model version page displays information about a specific registered model version... You can also perform tasks from this page

such as deleting a model version or transitioning a model version to another stage." This confirms the model version page is the correct location for this task.

Source: https://docs.databricks.com/en/mlflow/model-registry.html#model-version-page