
SageMaker Distribution, a pre-built Docker image with popular ML and data science packages, is now available as a first-party image in Amazon SageMaker Studio for streamlined machine learning development.


<div><p>This article announces the availability of SageMaker Distribution, a pre-built Docker image with popular ML and data science packages, now integrated into Amazon SageMaker Studio.</p><ul><li>Pre-built image includes PyTorch, TensorFlow, Keras, NumPy, scikit-learn, pandas, and JupyterLab</li><li>Available as first-party image in SageMaker Studio without custom image creation</li><li>SageMaker Python SDK now built-in with the image</li><li>Use @remote decorator to run local notebook code as remote SageMaker training jobs</li><li>Conda environment files available for reproducing environments locally</li><li>Customizable via direct package installation, Conda, or custom Docker images</li></ul><p>SageMaker Distribution simplifies ML development by providing pre-configured environments with common packages directly in SageMaker Studio, enabling seamless local-to-cloud workflow transitions.</p></div>


Related articles