Microsoft and Amazon Web Services have developed a new library named Gluon, which will allow developers of all levels to use artificial intelligence in their programs. This Gluon library is therefore mainly dedicated to the exploitation of Deep Learning. Indeed, designing Deep Learning models requires a lot of effort and great expertise, and Gluon’s goal is to facilitate the creation of neural networks and the training of models.
With the Gluon Library, developers can build machine learning models using a simple Python API and a range of prebuilt and optimized neural network components. Neural networks are built more easily using simple, concise code, without giving up performance. This requires large volumes of data and complex algorithms that have come to simplify and accelerate learning engines like Apache MXNet, Microsoft Cognitive Toolkit (CNTK), and Google TensorFlow.
To conclude, the Gluon interface created by Microsoft and Amazon Web Services makes it possible to create neural networks and training models as simple as creating an application. Gluon is actually democratizing Deep Learning by making this new interface accessible to all developers.
Gluon is available on GitHub: https://github.com/gluon-api/gluon-api/