Google Machine Intelligence research organization’s Google Brain Team just open-sourced TensorFlow. TensureFlow is a machine learning library used across Google to apply deep learning to different areas. It’s very good at perceptual and language understanding tasks – designed to facilitate research in machine learning, and to make it quick and easy to transition from research prototype to production system.
TensorFlow was originally developed by researchers and engineers at Google but the entire library was open-sourced yesterday. Not only was the code released under Apache 2.0 open source license, but Google also moved internal development efforts over to use a public repository so that day-to-day changes made by the team at Google are available to the public – this is HUGE!
This open-source release already supports machines (desktops/servers) and mobile devices and is already used in Google products. Standardizing a library and common toolset will help accelerate development in this field. My prediction is that in 3 years, we along with our customers and competitors will be using this in some form or the other.
But don’t ask me if I can build a robot butler by just importing this library – that will take a little longer.