Facebook & Microsoft introduce new open ecosystem for interchangeable AI frameworks – News


Image from Iconfinder

Artificial Intelligence (AI) developers, listen up. Facebook & Microsoft are today introducing Open Neural Network Exchange (ONNX) format, a standard for representing deep learning models that enables models to be transferred between frameworks.

According to a post on the FB blog, ONNX was the 1st step toward an open ecosystem where AI developers can easily move between state-of-the-art tools & choose the combination that is best for them.

When developing learning models, engineers & researchers have many AI frameworks to choose from. At the outset of a project, developers have to choose features & commit to a framework. Many times, the features chosen when experimenting during research & development are different than the features desired for shipping to production. Many organizations are left without a good way to bridge the gap between these operating modes & have resorted to a range of creative workarounds to cope, such as requiring researchers work in the production system or translating models by hand.

FB developed ONNX together with Microsoft to bridge this gap & “to empower AI developers to choose the framework that fits the current stage of their project & easily switch between frameworks as the project evolves.” Caffe2, PyTorch, & Cognitive Toolkit will all be releasing support for ONNX in September, which will allow models trained in one of these frameworks to be exported to another for inference, said the post.

For more, click here.

Don’t leave without reading this…

… we seek a favor. Online advertising revenues are falling, especially for media Sites. Also, as you may have realized, unlike many media & publishing organizations, we haven’t put up a paywall – we firmly believe that information, including our content must be accessible by all across the globe. But content takes a time & money to produce. If you’ve read the above piece & liked it, we urge you to contribute in our support.





•Share This•