ModelArts now supports the integration of Affinity Policy components when creating training jobs. This feature helps data scientists and AI engineers achieve more predictable performance and improved infrastructure efficiency.

A completely redesigned Training Job Creation page is now available.

• Modernized user interface.
• Simplified configuration workflow.
• Improved parameter organization and navigation.
• Faster and more intuitive training job setup.

The new experience reduces configuration complexity and helps users create and launch AI training tasks more efficiently.

Hardware specifications have been updated with a refreshed naming convention.
This update enhances transparency and simplifies resource selection for users.

Administrators can now modify the names of existing resource pools.
This enhancement helps organizations maintain clearer and more consistent resource management practices.

The tag management component has been refined and standardized.
These changes provide a cleaner and more intuitive user experience.

The frontend experience for Development Environment Public Pools has been optimized.
Users can now manage development environments more efficiently through a cleaner interface.

The Resource Management module has received a comprehensive set of usability improvements.

• Optimized terminology and interface text for improved clarity.
• Refined logical sub-pool management workflows.
• Improved process sequencing and operational guidance.
• Enhanced scale-up and scale-down experience.
• Optimized resource detail pages with clearer information presentation.

• Simplified resource administration.
• More efficient resource lifecycle management.
• Improved operational transparency.
• Better overall user experience for platform administrators and operators.

Further details on the service functionality are available in the ModelArts Help Center.