Amazon SageMaker AI Gets Serverless Fine-Tuning for NVIDIA Nemotron 3.5 Lightning

You can now customize NVIDIA Nemotron 3.5 Lightning directly in Amazon SageMaker AI using supervised fine-tuning, preference optimization, and reinforcement learning—all serverless, so you skip the infrastructure headaches and pay only for what you use. The feature rolls out across US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland), letting you adapt this efficient 3B-parameter model to your specific domains and workflows without breaking the bank.