Google is offering an open source, hardware-accelerated library for machine learning that runs in a browser. The library is currently supported only in the desktop version of Google Chrome, but the project is working to support more devices.
The Deeplearn.js library enables training of neural networks within a browser, requiring no software installation or back end. “A client-side ML library can be a platform for interactive explanations, for rapid prototyping and visualization, and even for offline computation,” Google researchers said. “And if nothing else, the browser is one of the world’s most popular programming platforms.”
[ Learn how to write apps that take full advantage of machine learning: Data in, intelligence out: Machine learning pipelines demystified • Google’s machine-learning cloud pipeline explained • R and Python drive SQL Server 2017 into machine learning . | Keep up with hot topics in programming with InfoWorld’s App Dev Report newsletter . ]
Deeplearn.js imitates the structure of the company’s TensorFlow machine intelligence library and NumPy, a scientific computing package based on Python. “We have also implemented versions of some of the most commonly used TensorFlow operations. With the release of Deeplearn.js, we will be providing tools to export weights from TensorFlow checkpoints, which will allow authors to import them into webpages for Deeplearn.js inference.”