Meta is handing over its PyTorch AI platform to the Linux Foundation
>
Meta has announced (opens in new tab) it is transferring its PyTorch artificial intelligence (AI) platform to the Linux Foundation.
Created by Facebook in 2016, PyTorch is an in-depth learning framework that has been used for many AI projects, with an emphasis on machine learning (ML) projects.
Ownership of the platform will now be taken over by the newly formed PyTorch Foundation of the Linux Foundation, whose board of directors includes AWS, AMD, Google Cloud, Meta, Microsoft and Nvidia, and says it will work. working closely with the project administrators and the developer community to support and focus on the ecosystem.”
PyTorch Linux Foundation
One of the great advantages of PyTorch is that developers can use Python as their main programming language. This is particularly beneficial given the flexibility and widespread adoption of this coding language.
In addition, because PyTorch uses a graphics processing unit accelerated framework, many developers who prefer GPUs to run ML models will benefit from it.
One such adoption of PyTorch’s machine learning is Facebook’s proprietary ‘M2M-100’ translation tool, which is capable of translating between any pair of 100 languages without relying on English translations.
Meta says the newly established PyTorch Foundation will focus on four key principles, namely: staying open; maintaining neutral branding; stay honest; and forging a strong technical identity.
Facebook’s parent company hopes not only that transferring ownership to an independent agency will help lay the groundwork for these four principles, but that the move will “[ensure] that decisions will be made in a transparent and open manner,” as the global tech giant comes under fire from all quarters, including questions about its algorithms.
Going forward, Meta says it will remain committed to and invest in the project, in addition to the support that PyTorch will receive from numerous other companies.