Databricks AI Runtime and NVIDIA make it easier to customize open models with enterprise data
Databricks AI Runtime (AIR) is now generally available, letting teams train and fine-tune models on serverless NVIDIA GPUs with NeMo and containers, alongside governed enterprise data and without managing GPU infrastructure. It pairs NVIDIA Kumo models for structured-data predictions and Nemotron Retriever models for finding evidence in unstructured data, with Genie Code, NVIDIA-specific Agent Skills and joint Nemotron cookbooks to help you get started.
* Databricks AI Runtime (AIR) is now Generally Available (GA), helping teams train and fine-tune models faster on serverless NVIDIA GPUs without the complexity of managing GPU infrastructure. * AIR supports NVIDIA NeMo and containers so teams can train and fine tune models on serverless NVIDIA GPUs alongside governed enterprise data. * Databricks uses NVIDIA Kumo open models to accelerate predictions from structured data and NVIDIA Nemotron Retriever models to find relevant evidence from unstructured data, enabling agents to make informed decisions and take action within Databricks workflows. * Developers can leverage Genie Code, NVIDIA-specific Agent Skills and our joint Nemotron cookbooks to get started with model customization on AIR.