Model Registry

Selecting a specific object in the Registry opens the Model Registry interface. This is the control panel for a specific artifact (model), providing access to detailed information and analysis tools. The LUML Model Registry is a centralized hub for organizing and tracking machine learning models throughout their lifecycle. It stores models, including the experiment run that generated them, and maintains a history of their versions. Users can annotate models with description and assign tags to support downstream workflows. This makes collaboration, governance, and model evolution easier to manage across teams.

Model Overview

The Overview tab acts as the technical passport of the object. It contains a structured dictionary of static metadata describing the model’s origin and characteristics. This includes:

• General Information - model name, creation date, file size.
• Manifest - a technical description of the .luml model contents.
• Tags - service metadata that adds searchable context to models for easier organization and workflow support.