Meta Cedes PyTorch to Linux Foundation
Meta AI this week announced that it is handing control of the popular open source PyTorch framework that it created to the newly set up PyTorch Foundation.
The PyTorch Foundation will operate as part of the nonprofit Linux Foundation, with a governing board composed of representatives from AMD, AWS, Google Cloud, Meta, Microsoft Azure, and Nvidia, with the intention of expanding over time.
PyTorch is an open-source machine learning framework based on the Torch library. As we reported previously, data scientists or AI researchers working on deep learning projects will probably use either TensorFlow or PyTorch.
The latter is favored by researchers, with most state-of-the-art (SOTA) models and publications of AI papers in PyTorch. This is due to its support for Python, as well as its support for tensors, which makes it possible to create and manipulate data in ways not possible using other frameworks.
Indeed, Meta AI claims more than 80 percent of researchers who submit their work at major ML conferences, such as NeurIPS or ICML, harness PyTorch. And since it was created in 2016, some 2,400 contributors have built more than 150,000 projects on the framework, according to Meta AI.
Vendor-neutral framework
The creation of the PyTorch Foundation ensures that decisions about PyTorch will be made in a transparent and open manner by a diverse group of members. The foundation will prioritize the continued growth of the PyTorch community, which is credited as the real driving force behind the project’s success.
The transition of PyTorch to the PyTorch Foundation will not affect the existing PyTorch code or its technical governance structure. Moving forward, releases, features, and technical direction will still be driven by the maintainers, committers, and contributors to the project.
“Meta has worked steadily to nurture the community-driven growth that has fueled PyTorch’s success. We’ve committed hundreds of engineers to the framework and strongly supported product development and community outreach. But the time is right for a new home for PyTorch,” wrote the Meta AI team in a blog post.
“The PyTorch Foundation will strive to adhere to four principles: remaining open, maintaining neutral branding, staying fair, and forging a strong technical identity. One of the foundation’s main priorities is to maintain a clear separation between the business and technical governance of PyTorch.”
Image credit: iStockphoto/Flashvector
