Amazon has decided to follow in the footsteps of Google and other technology companies and open source its deep learning software.
The company has released its deep learning software DSSTNE (pronounced Destiny) on GitHub under the open source Apache license. Deep learning has received a lot of attention in recent months, and many technology companies are now developing their own software to help teach computers.
About five months ago, Google also open sourced its own deep learning library, TensorFlow. Amazon has posted a FAQ page and library to explain why they decided to open source DSSTNE. “We are releasing DSSTNE as open source software so that the potential 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 work together to improve it. But more importantly, we hope that DSSTNE will foster innovation in more areas.”
Amazon’s deep learning software has limitations and currently cannot support convolutional workloads for image recognition and has limited support for recurrent neural networks. However, this software can utilize two graphics processing units (GPUs) at the same time, which other frameworks cannot do.
DSSTNE shows some performance advantages over Google’s TensorFlow with Amazon, with a 2.1x speedup compared to competing software on g2.8xlarge GPU instances in the Amazon Web Services (AWS) public cloud.
Deep learning and machine learning are the directions the industry is moving in, and with companies like Amazon, Google, and others making their software available through open source licenses, these technologies will be developed at an even faster pace.
Published under license by ITProPortal.com, a publication of Net Communities Ltd. Unauthorized reproduction is prohibited.
Photo credits: rvlsoft/shutterstock
