AWS (Amazon Web Service), subsidiary of Amazon.com, has become the newest member of deep learning community’s alliance on the ONNX (Open Neutral Network Exchange), and recently came out with advanced artificial intelligence environment.
Facebook and Microsoft led the effort of the framework.
Part of the collaboration, AWS made it’s own open source Python Package, and other side ONNX-Mxnet already available as a deep learning framework that offers application programming interfaces across multiple language including Scala, Python and open source static software R.
The ONNX design will help the developers to build and train models for other frameworks, like PyTorch, Microsoft Cognitive Toolkit or Caffe2, AWS Deep Learning Engineering Manager, Hagay Lupesko and Software Developer Roshani Nagmote wrote in an online post last week. It will let developers import those models into MXNet, and run them for inference.
Microsoft and Facebook launched support for ONNX this summer, a shared model of interoperablity for the enhancement of the AI. Microsoft committed its Cognitive Toolkit, Caffe2 and PyTorch to support ONNX.
“Cognitive Toolkit and other frameworks make it easier for developers to construct and run computational graphs that represent neural networks,” Microsoft said.
Inceptive versions of ONNX code and documentation were made available on Github.
AWS and Microsoft Introducing Gluon a new library for machine learning, a new interface in Apache MXnet that helps developer to build and train deep learning modules.
According to Apache MXnet community earlier released a version 0.12 of MXnet, which extends Gluon functionality for it’s new research,
“This looks like a great way to deliver inference regardless of which framework generated a model,” said Paul Teich, principal analyst at Tirias Research.
“This is basically a framework-neutral way to deliver inference,” he said LinuxInsider.
AWS and Microsoft and others cloud providers are under pressure from customer to be able to train on one network while dispatching on another, in order to advance AI, Teich Said.
“Framework interoperability is a good thing, and this will only help developers in making sure that models that they build on MXNet or Caffe or CNTK are interoperable,” Tractica’s Kaul quote.
These companies like ONNX, AWS, Facebook, Microsoft they’ll know that they owe a significant amount of their triumph to the software development and open source, quoted Jeff Kaplan.
“The Open Neural Network Exchange is committed to producing similar benefits and innovations in AI,” he told LinuxInsider.
Numerous major tech companies have announced their plans to use and speed up the development of AI collaboration, In order to create uniform framework for development and research.