Amazon open sources its deep learning software

The Deep Scalable Sparse Tensor Network Engine (DSSTNE) library is now available under the Apache 2.0 license.
Written by Stephanie Condon, Senior Writer

Following the lead of other tech giants, Amazon has open sourced its deep learning software. The Deep Scalable Sparse Tensor Network Engine (DSSTNE) library is now available on GitHub under the Apache 2.0 license.

Amazon explained on its FAQ page that DSSTNE, pronounced "Destiny," helps Amazon's hundreds of millions of customers shop. "We help them discover the right product from an immense catalog of products," the company explained. "Making good recommendations requires neural networks."

The web giant said it's releasing DSSTNE as open source software "so that the promise of deep learning can extend beyond speech and language understanding and object recognition to other areas such as search and recommendations. We hope that researchers around the world can collaborate to improve it. But more importantly, we hope that it spurs innovation in many more areas."

DSSTNE is distinct from other deep learning libraries because it's designed to support problems with sparse data and is faster at tackling them, Amazon says. In fact, the company claims DSSTNE is 2.1 times faster than TensorFlow, Google's open source machine learning system, on a g2.8xlarge GPU instance in the Amazon Web Services cloud.

Google open sourced TensorFlow back in November, while Facebook took its own machine learning and artificial intelligence tools to the open source community in early 2015. Meanwhile, in December Amazon Web Services joined several other major players in investing in an open-source, nonprofit AI research project called OpenAI.

Editorial standards