Amazon SageMaker AI now supports serverless model customization for NVIDIA Nemotron 3.5 Lightning

Why it matters

You can now fine-tune NVIDIA's 30B-parameter model on SageMaker without managing infrastructure, reducing time to deploy custom AI.

Ask your team

Should we use SageMaker to customize Nemotron 3.5 Lightning for our domain instead of self-hosting?

What happened

Amazon SageMaker AI now supports serverless model customization for NVIDIA Nemotron 3.5 Lightning using SFT, DPO, and RFT. The model has 3B active parameters and 30B total parameters with a hybrid Mixture-of-Experts architecture.

From the source

“5 Lightning model using supervised fine-tuning (SFT), Direct Preference Optimization (DPO), and reinforcement fine-tuning (RFT). This is one of the latest open-weight models from NVIDIA and employs a hybrid Mixture-of-Experts architecture with 3B active parameters and 30B parameters in total.”