Amazon SageMaker JumpStart now provides granular access control for foundation models
Starting today, enterprise admins using Amazon SageMaker JumpStart can easily configure granular access control for foundation models (FM) that are discoverable and accessible to users within their organization. Amazon SageMaker JumpStart is a machine learning (ML) hub that offers pretrained models and built-in algorithms to help you quickly get started with ML.
Amazon SageMaker JumpStart provides access to hundreds of FMs, however many enterprise admins want more control over the FMs that can be discovered and used by users within their organization (e.g., only allowing Apache 2.0 license models to be discovered). With this new feature, enterprise admins can now create private hubs in SageMaker JumpStart through the SageMaker SDK, and add specific FMs into private hubs that can be accessible to users within their organization. Enterprise admins can also set up multiple private hubs that are tailored for different roles or accounts with a different set of models. Once set, users will be able to view hubs and models they are allowed to view and use through SageMaker Studio and the SageMaker SDK.
Granular control of FMs in SageMaker JumpStart can be used initially in US-East (Ohio) starting today. To learn more, see the blog and product page.