Registering models for tracking with AI Factsheets
You can track models from a project, space, or catalog so that the details for the model lifecycle display in a model use case, stored in a model inventory. After you enable model tracking, details for the model are recorded in the associated factsheet as part of your AI governance strategy.
Tracking a model
When a model is tracked, the AI Factsheets service creates a relationship between the model and the associated model use case. All of the facts regarding how the model was created and where the model is on the path to production are recorded in the use case for ease of review and governance.
Tracking a model from the model details page of a project, space, or catalog
Click a model name to view the model details in a project, space, or catalog.
Click Track this model.
Choose to associate the model with an existing model use case or create a new use case.
When you save the tracking information, a link to the model inventory is added to the model details page. Click the link to open the model use case in the inventory.
Optionally update the model use case. For example, add tags, supporting documentation, or other details.
When you view the model detail page, tracking information is displayed.
Registering a model in a notebook
If your model is in a notebook, you can insert code similar to this sample code at the end of the notebook to register the model with a model use case:
model = facts_client.assets.get_model(model_id=<model id>,container_type="space or project",container_id=<space id or project id>) model.add_tracking_model_usecase(model_usecase_name=<name>,model_usecase_desc=<description>,model_usecase_catalog_id=<catalog id where model use case should be created>)
Removing a model from an use case (untracking)
Click Untrack to disassociate the model from the model use case.
Learn about viewing model use cases and factsheets.
Parent topic: Managing an inventory of model assets.