AWS and Microsoft may be arch rivals when it comes to competing for business in cloud storage and services, but when it comes to breaking ground in newer areas where volumes of data make a difference to how well the services work and creating systems that are easier to use, collaboration is key. Today's announcement is one such project, with both firms pooling their resources to push the boundaries of machine learning development.
"With the Gluon interface, developers can build machine learning models using a simple Python API and a range of prebuilt, optimized neural network components".
Thanks to Gluon, developers should expect a concise and easy to understand program interface, the ability to quickly prototype and experiment with neural network models, and a training method that has minimal impact on the speed of the underlying engine. It's then implemented using code structures similar to the ones used in app and website development. And because of the complexity of each component, the process can take a long time.
CVP of Microsoft AI and Research, Eric Boyd, believes that the Gluon Interface will provide developers with a "freedom of choice". Conversely, some engines focus on offering simpler model building at the expense of training performance.
Vidal Retires From Chile National Team After World Cup Failure
The Chileans finished level on points with Peru, but trailed on goal difference to end up in sixth place in the final standing. Also Chile's Argentinean manager Juan Antonio Pizzi announced his resignation from the team after Tuesday's disappointment.
The problem with the construction of neural networks seems to lie in maintaining balance between model building and training performance. Additionally, it will allow them to debug and update neural networks much more smoothly. The end product is a powerful deep learning interface that's both accessible and performant.
Essentially, the two companies built Gluon "so building neural networks and training models can be as easy as building an app", Swami Sivasubramanian, vice president of Amazon AI, said in the release. Thanks to a jointly developed reference specification, Gluon will also work on any deep learning engine, with support for Apache MXNet available today and support for Microsoft Cognitive Toolkit coming soon.
Cloud services rivals Amazon Web Services and Microsoft are collaborating on a deep learning library created to accelerate development of models used to construct neural networks. Integrating with Apache MXNet and Microsoft Cognitive Services is a start but for now teams using architecture from Google or Amazon are unable to access Gluon's benefits.