Amazon SageMaker JumpStart provides a feature called private hubs, allowing organizations to share and manage access to pre-trained models and solutions within their enterprise. Admins can configure granular access controls, enabling tailored model repositories for different teams and use cases. Users can access these models via SageMaker Studio and the Python SDK. The feature supports cross-account sharing using AWS Resource Access Manager (AWS RAM), facilitating collaboration while maintaining control. The private hubs decouple model curation from consumption, aligning AI initiatives with corporate policies and regulatory requirements.

10m read timeFrom aws.amazon.com
Post cover image
Table of contents
Solution overviewPrerequisitesCreate a private hub, curate models, and configure access control (admins)Interact with allowlisted models (users)Cross-account sharing of private hubsConclusionAbout the Authors

Sort: