Amazon SageMaker notebooks

Fully managed notebooks in JupyterLab for exploring data and building ML models

What is SageMaker notebooks?

Launch fully managed JupyterLab from Amazon SageMaker Studio in seconds. Use the integrated development environment (IDE) for notebooks, code, and data. You can use the quick start, collaborative notebooks in the IDE to access purpose-built ML tools in SageMaker and other AWS services for your complete ML development, from preparing data at petabyte scale using Spark on Amazon EMR, to training and debugging models, deploying and monitoring models and managing pipelines – all in one web-based visual interface. Easily dial compute resources up or down without interrupting your work.

Benefits of SageMaker notebooks

Launch fully managed JupyterLab in seconds in Amazon SageMaker Studio. SageMaker Studio comes preconfigured with the SageMaker distribution containing popular packaging for ML, including deep learning frameworks, such as PyTorch, TensorFlow, and Keras, and popular Python packages, such as NumPy, scikit-learn, and pandas.
Scale your compute resources up or down with the broadest selection of compute-optimized and GPU-accelerated instances in the cloud.
Use the generative AI–powered coding companion and security tools to write high-quality code faster. Generate, debug, and explain the source code with Amazon CodeWhisperer and conduct security and code quality scans with Amazon CodeGuru.
Build unified analytics and ML workflows in the same notebook. Run interactive Spark jobs on Amazon EMR and AWS Glue serverless infrastructure, right from your notebook. Monitor and debug jobs faster using the inline Spark UI. Easily automate your data prep by scheduling the notebook as a job in few simple steps.

Build ML at scale

Quick start

Launch fully managed JupyterLab in Studio in seconds. SageMaker Studio comes pre-configured with pre-built SageMaker distribution containing popular packaging for ML, including deep learning frameworks like PyTorch, TensorFlow, and Keras; popular Python packages like NumPy, scikit-learn, and panda to help you get started with model building.

Elastic compute

Scale your underlying compute resources up or down, and use shared persistent storage to switch compute, all without interrupting your work. Pick from the broadest selection of compute resources offered by AWS, including the most powerful GPU instances for ML.

Boost ML development productivity

Data preparation

Simplify your data workflows with a unified environment. Create, browse, and connect to Amazon EMR clusters and AWS Glue Interactive Sessions directly from JupyterLab. Use the built-in data preparation capability to visualize data and improve data quality.

Notebook jobs

You can use SageMaker notebook jobs to create a non-interactive job to either run on demand or on a schedule. Use an intuitive user interface or SageMaker Python SDK to schedule your jobs right from JupyterLab. Once a notebook is selected, SageMaker notebook takes a snapshot of the entire notebook, packages its dependencies in a container, builds the infrastructure, runs the notebook as an automated job on a schedule set by the practitioner, and deprovisions the infrastructure upon job completion. SageMaker notebook jobs is also available as a native step in Amazon SageMaker pipelines to enable you to automate your notebooks into multi-step workflows with dependencies for CI/CD deployment within a few lines of code.

AI-powered tools

Amazon Q Developer provides ‘how-to’ guidance on SageMaker features, code generation assistance, and support for troubleshooting in the JupyterLab environment. Simply ask your questions in natural language, such as "How do I deploy my model on a SageMaker endpoint for real-time inference?", and Amazon Q Developer will provide step-by-step instructions and code to get you started. When you encounter errors while executing the code, the Amazon Q Developer is there to lend a helping hand. Just ask it to fix the error, and it will provide detailed steps to debug and resolve the issue.