Goal: 

RHEL AI builds 2 versions of the bundled granite model. 
1. OCI Artifact image eg: here

2. OCI modlcar Image eg: here

We in Quay.io (and Quay) want to display the metdata from the image and in the future make the modelcard metadata searchable via API. For this we need to parse the model card. 

In order to identify which image is an AI image and which layer has the model card we rely on annotation on the image manifest. This works for the OCI Artifacts but there are no annotations in the modelcar version of the image. Can you please add them when building model car images as well 

Acceptance Criteria:

• Model car images have the same annotations as the OCI artifact images
• The model Card is in a single layer which is annotated the same as the OCI artifact image
• (optional) A tool like olot takes care of this. This can be utilized to generate the model car images