Amazon SageMaker AI now supports serverless model customization for NVIDIA Nemotron 3.5 Lightning
Amazon SageMaker AI supports serverless model customization for the NVIDIA Nemotron 3.5 Lightning model, enabling SFT, DPO, and RFT customization without cluster management.
Amazon SageMaker AI supports serverless model customization for the NVIDIA Nemotron 3.5 Lightning model, NVIDIA's latest open-weight model with a hybrid Mixture-of-Experts architecture using 3B active and 30B total parameters. Customization allows fine-tuning the foundation model on proprietary data to maintain state-of-the-art quality with smaller, suitable models, reducing cost and latency. Use SFT with labeled data to improve domain-specific task accuracy, DPO to align outputs with organizational tone, and RFT with reward signals to enhance new task performance. Serverless customization lets SageMaker AI handle all infrastructure provisioning and training orchestration, enabling focus on data and evaluation instead of cluster management, with pay-as-you-go pricing. It is available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland). Start customization jobs from the Amazon SageMaker Studio model page or via the SageMaker Python SDK; see Amazon SageMaker AI model customization documentation for details.
Why it matters
This update is useful for developers and data scientists who need to customize AI models for specific domains or tasks. SageMaker AI's serverless capabilities reduce infrastructure management overhead and improve cost efficiency.