Amazon SageMaker AI announces training session APIs for serverless model customization
Amazon SageMaker ยท 2026-10-22
Actions
Technical Details
| Regions | us-east-1, us-west-2, ap-south-1 |
|---|---|
| Cost Impact | Neutral |
What This Means
For DevOps Teams
Integrate the new training session APIs into your AI workflows to enable custom training loops and techniques, reducing the need for manual GPU management and allowing for more efficient and flexible model training processes.
For Platform Teams
Adopt the new training session APIs in Amazon SageMaker AI to provide your teams with advanced, customizable AI training capabilities, enhancing your platform's AI/ML offerings and reducing operational toil associated with GPU management.
For Executives
Evaluate the new training session APIs in Amazon SageMaker AI to enhance your AI model training capabilities, allowing for custom training loops and techniques without managing GPU infrastructure, ultimately driving innovation and competitive advantage in your AI strategy.
Source
Related Amazon SageMaker Updates
- Announcing new ways to customize models with SageMaker AI, without managing GPUs (2026-10-22)
- Amazon SageMaker AI now supports serverless model customization for Muse Glimmer, Qwen 3.8 and Ministral 3 models (2026-10-22)
- Beam model by Reflection is now available on Amazon SageMaker JumpStart and AWS Trainium (2026-10-21)
- CVE-2026-112408 - Missing bucket ownership verification in Amazon SageMaker Python SDK default_bucket() (2026-10-21)
- Amazon SageMaker AI now supports custom sequence lengths for serverless model customization jobs (2026-10-21)