Serverless GPU Python API
Warning
serverless_gpu is deprecated and no longer receives new features. Starting in GPU environment
version 6, the Databricks AI Runtime SDK
(databricks.air) replaces it. GPU environment version 6 includes both packages, and version 7
and above include only the Databricks AI Runtime SDK. To migrate, see
Migrate from serverless_gpu.
Serverless GPU API is a lightweight, easy-to-use library that simplifies running multi-GPU workloads on Databricks AI Runtime, directly from Databricks notebooks.
It provides the same functionality as running torchrun train.py or accelerate launch train.py
from the command line, but integrates seamlessly into the Databricks notebook environment.
For concepts, usage guides, and examples, see the AI Runtime documentation.