PyTorch on Azure

Simple and seamless PyTorch experience in the cloud

PyTorch is an open-source deep learning framework that provides a seamless path from research to production. As a Python-first framework, PyTorch enables you to get started quickly, with minimal learning, using your favourite Python libraries.

Azure supports PyTorch across a variety of AI platform services. Whether you are getting started with PyTorch, training a model or deploying models to production, Azure helps you accelerate your project from the comfort of your own coding environment.

Three ways to use Azure for PyTorch development

Initiate your projects with Azure Notebooks

Get started quickly with a free, web-based Jupyter Notebooks platform that comes with PyTorch preinstalled. Begin experimenting right away with support from our library of official PyTorch tutorials that you can easily clone into your own library.

Get started

Develop with preconfigured Data Science Virtual Machines

Go straight to development with custom Windows or Linux virtual machines specially configured for machine learning workloads. Data Science Virtual Machines come preinstalled with PyTorch as well as the necessary GPU drivers and a comprehensive suite of other popular data science tools. You get a frictionless development experience out of the box and the ability to integrate with all Azure hardware configurations, from GPUs to FPGAs.

Learn More

Accelerate your workflow with Azure Machine Learning service

Train and deploy PyTorch models with ease from your preferred Python environment, such as Jupyter Notebooks, Azure Notebooks and Visual Studio Code, with the Azure Machine Learning service Python SDK. Azure Machine Learning service not only removes the heavy lifting of end-to-end machine learning workflows, but also handles housekeeping tasks like data preparation and experiment tracking, cutting time to production from weeks to hours.

Learn More

Why Pytorch?

Python-first

PyTorch is deeply integrated with Python so you can use your favourite Python libraries, packages and debuggers. This allows for rapid prototyping and development of PyTorch models.

Hybrid front-end

Enjoy ease-of-use and flexibility for development, while enabling speed, optimisation and functionality in C++ runtime environments.

Native ONNX support

PyTorch supports native export of models in the standard ONNX (Open Neural Network Exchange) format. This facilitates interoperability with ONNX-compatible frameworks and inferencing on a variety of hardware platforms and runtimes, including the open-source ONNX Runtime.

Active community

Using PyTorch, you join a highly supportive community of researchers and engineers developing rich libraries and tools in areas like computer vision, natural language processing and reinforcement learning. This network can provide an invaluable resource for technical education and guidance.

Related products and services

Azure Machine Learning service

Build models rapidly and operationalise at scale from cloud to edge

Data Science Virtual Machines

Go straight to development with a custom virtual machine, preconfigured for machine learning workloads

Azure Notebooks

Get started quickly with a free, web-based Jupyter Notebooks platform

Accelerate your PyTorch project in the cloud with Azure