Amazon has become the latest company to join ONNX, a new open source ecosystem for interchangeable AI models that Microsoft and Facebook launched this past September.
This move by AWS is a massive step towards increasing interoperability in the AI tools community.
The ONNX open ecosystem is supported by a wide array of frameworks and hardware vendors including industry leaders like AMD, ARM, Huawei, IBM, Intel and Qualcomm.
AWS is enabling support for the technology for its Apache MXNet deep learning framework.
As the cloud giant explained, this development is a significant stamp of approval to the concept of open data models, in an excitingly dynamic field. Advanced research into deep learning has been going on for several years now, but despite this focus, there are not lot of methods for building data models.
What’s more, it is also fairly difficult for developers to switch frameworks upon discovering that their model may benefit from a different approach.
This is the idea behind ONNX, which is designed to serve as a bridge between open source deep learning frameworks like Caffe2 that Facebook developed, and Cognitive Toolkit, developed at Microsoft. It now also supports Apache MXNet, thanks to this support by AWS.
Overall, this is a very positive development for those at the forefront of artificial intelligence research, and good things usually come out when major players get together for a common goal.
Check out more details on ONNX-MXNet here on its GitHub page.
0 comments