PyTorch on AWS

A highly performant, scalable, and enterprise-ready PyTorch experience on AWS

Accelerate time to train with Amazon EC2 instances, Amazon SageMaker, and PyTorch libraries.

Speed up research prototyping to production scale deployments using PyTorch libraries.

Build your ML model using fully managed or self-managed AWS machine learning (ML) services.

How it works

PyTorch on AWS is an open-source deep learning (DL) framework that accelerates the process from ML research to model deployment.
Diagram showing how you can train models in PyTorch with a TorchServe model server.

Use cases

Distributed training for large language models

Use PyTorch Distributed Data Parallel (DDP) systems to train large language models with billions of parameters.

Learn more »

Inference at scale


Scale inference using SageMaker and Amazon EC2 Inf1 instances to meet your latency, throughput, and cost requirements.

Learn more »

Multimodal ML models


Use PyTorch multimodal libraries to build custom models for use cases such as real-time handwriting recognition.

Learn more »

How to get started

Get started with PyTorch on AWS

Find everything you need to get started with PyTorch on AWS.

Learn more »

Explore more about PyTorch on AWS

Check out key features and capabilities to begin working with PyTorch.

Explore PyTorch »

Build with Amazon SageMaker JumpStart

Discover prebuilt ML solutions that you can deploy with a few clicks.

Learn more »

Explore more of AWS