New features now available in GA include the ability to… place customized tags on your workspaces and computes, search for machine learning assets across all workspaces, resource groups, and subscriptions within your organization, secure the ingress and egress of managed online endpoints without needing additional configuration, build custom metrics views, and simplify data security by making it easier for enterprises to use sensitive data for model training.
Add tags to workspaces and computes: You can now add, view, update, and/or delete customized tags on your workspaces and computes to gain deeper insights into cost patterns, spend patterns, and governance scenarios.
Search for Azure Machine Learning assets across multiple workspaces: You can now create structured queries to quickly find machine learning assets, view all workspaces in a new global homepage, and collaborate beyond the workspace boundary.
Manage network isolation easily for Managed Online Endpoints: You can now secure the ingress and egress of managed online endpoints to ensure compliance with enterprise security standards.
Customize your view of job metrics: You can now customize the view of metrics charts and resize/rearrange layouts to better understand and analyze your experimentation results.
Compartmentalize access to data with identity passthrough:You can now grant or deny user access to specific data using role-based access controls (RBAC). Additionally, when you submit a training job in identity passthrough mode, Azure Machine Learning uses that identity to authenticate against data storage.