sagemaker-python-sdk

A library for training and deploying machine learning models on Amazon SageMaker

What is sagemaker-python-sdk?

A library for training and deploying machine learning models on Amazon SageMaker

Overview of sagemaker-python-sdk

What it is

sagemaker-python-sdk is an open source library for training and deploying machine learning models on Amazon SageMaker. It is licensed under Apache-2.0 and is available as open source software. The SDK supports training and deployment using popular deep learning frameworks Apache MXNet and PyTorch, as well as Amazon algorithms which are scalable implementations of core machine learning algorithms optimized for SageMaker and GPU training. Users can also train and host models using their own algorithms built into SageMaker compatible Docker containers. Version 3.0.0 introduces a modernized architecture with separate PyPI packages for core, training, serving, and MLOps capabilities. The SDK is tested on Python 3.10, 3.11, and 3.12, and supports Unix/Linux and Mac operating systems. The library includes telemetry enabled by default, which can be opted out by setting the TelemetryOptOut parameter to true in the SDK defaults configuration.

Key features

  • Supports training and deployment with Apache MXNet, PyTorch, Amazon algorithms, and custom algorithms in SageMaker compatible Docker containers.
  • Version 3.x introduces a modular architecture with separate PyPI packages: sagemaker-core, sagemaker-train, sagemaker-serve, and sagemaker-mlops.
  • Unified training via the ModelTrainer class, replacing the V2 Estimator and framework-specific classes.
  • Unified inference via the ModelBuilder class, replacing the V2 Model and framework-specific classes.
  • Object-oriented API with auto-generated configs aligned with AWS APIs.
  • Supports hyperparameter tuning, model registry, processing jobs, transform jobs, and SageMaker Clarify.
  • Telemetry is included to track usage of SageMaker functions; opt-out is available.

Use cases

Training and deploying machine learning models on Amazon SageMaker using deep learning frameworks or Amazon algorithms. Running custom training containers with data from S3. Deploying trained models for real-time inference. Hyperparameter tuning jobs. Model registry and MLOps pipelines. Processing jobs with frameworks like PyTorch and Scikit-learn. Transform jobs for batch inference. Using SageMaker Clarify for bias detection and explainability.

Getting started

To install the latest version of SageMaker Python SDK 3.x:pip install --upgrade sagemakerTo install from source:git clone https://github.com/aws/sagemaker-python-sdk.gitcd sagemaker-python-sdkpip install .For detailed instructions, see the repository README.

License and pricing

The license is Apache-2.0. The pricing type is open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/awssagemaker-python-sdk.svg)](https://yunodb.com/tools/awssagemaker-python-sdk?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/awssagemaker-python-sdk?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/awssagemaker-python-sdk.svg" alt="Featured on YunoDB" height="20"></a>