Key Highlights:
- The program will allow service providers to develop and provide customized enterprise-grade support to their customers
- The support program was developed in response to feedback directly from PyTorch users
- The launch comes after Microsoft expanded its participation in PyTorch
PyTorch Enterprise Support Program
In collaboration with Microsoft, Facebook has created and launched the PyTorch Enterprise Support Program which allows service providers to develop and provide customized enterprise-grade support to their customers. According to reports by Facebook, the PyTorch Support Program was created in response to feedback directly from PyTorch users that are developing models in the production environment of mission-critical apps.
PyTorch Enterprise Support Program
According to Facebook, the PyTorch Enterprise Support Program can be availed by all service providers. It is specifically designed to offer sharing and improving PyTorch’s Long-Term Support (LTS) along with mutual benefit of all program participants. To benefit the open source community, hotfixes developed by the participants are scrutinized and regularly fed back to PyTorch’s LTS release through PyTorch’s pull request process.
With the PyTorch Enterprise Support Program, the standard way of researching and implementing will not change with different releases of PyTorch. In order to participate, service providers must apply and meet a set of program terms and certification requirements. After approval, the service provider can become a program participant and offer a packaged PyTorch Enterprise support service with LTS, prioritized troubleshooting, useful integrations, and much more.
PyTorch Open Source
PyTorch was officially launched by Facebook in January 2017. PyTorch is an open source machine learning library that is based on the scientific computing framework Torch. Torch is a scripting language that is based on the Lua programming language. Despite the availability of TensorFlow since 2015, Pytorch continues to prevail and grow rapidly amongst the data science and developer communities. In 2020, PyTorch claimed one of the top spots for a fast-growing open source project. As of 2019, the number of contributors on the platform increased by more than 50% year-on-year to nearly 1,200.
Microsoft’s partnership with Facebook
In 2017, Microsoft worked with Facebook to create ONNX for faster reasoning and to promote interoperability between AI hardware such as semiconductors and software such as machine learning frameworks. The launch of the PyTorch Enterprise Support Program comes after Microsoft expanded its participation in PyTorch to take ownership of the development and maintenance of the PyTorch build for Windows.
The aim is to bring and match the experience on Windows in line with other platforms, like Linux. As per past records, PyTorch on Windows has lagged behind due to a lack of test coverage, a convoluted installation experience, and missing functionality.
Also Read: Microsoft Develops Programming Tool for Low-Level Codes
