-
Epic
-
Resolution: Unresolved
-
Blocker
-
None
-
None
-
quay-ai-model-card-render
-
False
-
None
-
False
-
Not Selected
-
In Progress
-
PROJQUAY-8253 - Integration with OpenShift AI Model Registry in Quay
-
13% To Do, 38% In Progress, 50% Done
Epic Goal
- The primary objective of this epic is to integrate a feature that enables the rendering of model cards directly within the Quay user interface for any machine learning models stored in our repository. This enhancement will provide users with comprehensive insights into each model, enhancing their understanding and usability of the stored models.
Why is this important?
- Customer Satisfaction: Customers have expressed a need for this feature to better understand the capabilities and limitations of the machine learning models they store in Quay. This will increase customer satisfaction and potentially lead to more extensive use of our platform.
- Transparency and Trust: By providing model cards, we promote transparency regarding the training methodology, performance metrics, and intended usage scenarios for each model. This transparency can foster trust amongst users as they make informed decisions about which models to utilize.
- Improved Model Utilization: Detailed insights provided by model cards will aid users in making optimal choices when selecting a suitable model for their specific tasks or projects. This could lead to increased model utilization and overall efficiency.
Scenarios
- Scenario 1: A data scientist logs into Quay's UI, navigates to the repository containing a model. The system displays the associated model card in the user interface.
- Scenario 2: A team of engineers needs to assess different models for their project. They access Quay's UI and are able to view model cards for each of the models in the repository.
- Scenario 3: A new user joins an organization that uses Quay for machine learning models. They delete a model in the repository and Quay automatically deletes the associated model card.
- Scenario 4: A user pushes a new version of an existing model to Quay with an updated model card. The Quay UI defaults to showing the latest model card, but allows the user to view the model card for the old version if requested.
Acceptance Criteria
- All Scenarios are passing
Dependencies (internal and external)
- ...
Previous Work (Optional):
- …
Open questions::
- How will Quay retrieve the model card from the image?
- How should the UI be designed to show one/many model cards?
Done Checklist
- CI - CI is running, tests are automated and merged.
- Release Enablement <link to Feature Enablement Presentation>
- DEV - Upstream code and tests merged: <link to meaningful PR or GitHub Issue>
- DEV - Upstream documentation merged: <link to meaningful PR or GitHub Issue>
- DEV - Downstream build attached to advisory: <link to errata>
- QE - Test plans in Polarion: <link or reference to Polarion>
- QE - Automated tests merged: <link or reference to automated tests>
- DOC - Downstream documentation merged: <link to meaningful PR>
- is blocked by
-
RHELAI-3533 Build RHEL AI Model Car images with annotations to identify AI images in quay.io/quay
-
- In Progress
-
- is incorporated by
-
OCPSTRAT-1692 AI Workloads for OpenShift
-
- In Progress
-
- links to
-
RHBA-2025:2467 Red Hat Quay v3.14.0 bug fix release
- mentioned on