Microsoft has now added enterprise support for PyTorch AI on Azure. With this, Microsoft aims to give users of PyTorch a more reliable production experience.

For the uninitiated, PyTorch is a library for the Python programming language that integrates with important Python add-ons like NumPy and data-science tasks that require faster GPU processing. 

“This new enterprise-level offering by Microsoft closes an important gap. PyTorch gives our researchers unprecedented flexibility in designing their models and running their experiments,” Jeremy Jancsary, a senior principal research scientist at Nuance, said.

Notably, last year Microsoft took over the Windows 10 PyTorch library from Facebook to boost GPU-accelerated machine learning training on Windows 10’s Subsystem for Linux (WSL). The idea was to take a bigger role in the maintenance of PyTorch for Windows. “The direct involvement of Microsoft lets us deploy new versions of PyTorch to Azure with confidence,” he said. 

The latest move is expected to give Microsoft Premier and Unified Support for Enterprise customers more benefits, such as prioritized requests, hands-on support, and solutions for hotfixes, bugs and security patches. Additionally, Microsoft has promised to provide long-term support (LTS) to selected versions of PyTorch for up to two years, enabling a stable production experience without frequent major upgrade investment.

“Microsoft Premier support customers are automatically eligible for PyTorch Enterprise. They can request hotfixes directly to the dedicated Azure team. The requests will be prioritized, quickly addressed, and deployed in the next version of PyTorch. The latest release of PyTorch will be integrated with Azure Machine Learning, along with other PyTorch add-ons including ONNX Runtime for faster inferencing,” said Microsoft in a statement. 

All hotfixes are fed back to the LTS PyTorch distribution so everyone in the PyTorch community can use them. “Train and deploy models reliably at scale using a built-in PyTorch environment within Azure Machine Learning, and rest assured that your entire PyTorch stack is fully supported through PyTorch Enterprise,” the company said.