Google has finally given a BIG upgrade to its open source AI – Tensorflow
Google’s open source machine learning TensorFlow, which became the most popular project on GitHub, has added another feather to its cap with its new multi-server version of TensorFlow.
The most notable feature of this tool, which many TensorFlow users have been asking for since the tool made its public debut in late 2015, is its ability to operate on multiple devices at the same time.
TensorFlow is based on a branch of Artificial Intelligence called deep learning, which draws its inspiration from the way, how human brain cells communicate with each other. In other words, TensorFlow can be used to teach computers how to process data in the same way how the human brain handles information.
Also, deep learning has become central to the machine learning efforts of tech giants such as Microsoft, Facebook, and Yahoo, all of which are already busy in releasing their own AI projects.
According to Google’s chief executive, Sundar Pichai – It’s up to 5 times faster than its predecessor. Now, this explains the delay in releasing its more powerful new version. Also, Google was facing difficulties in adapting the software to be used outside Google’s highly customized data centers.
Distributed TensorFlow is powered by gRPC library, which supports training on hundreds of machines in parallel. It also complements Google’s recent announcement of Google Cloud Machine Learning, which enables the user to serve its open source tool models using Google Cloud Platform.
TensorFlow was developed to improve services Google users use on a regular basis. It has allowed the photo tool to identify many of the subjects in the images uploaded to its servers and taught the translation app to understand more of language’s peculiarity.
It has also made it easier for Google’s mobile apps to understand what people say when giving verbal instructions to its search engine.
Well, not everyone needs to run this open source machine learning on hundreds or thousands of servers. But many startups and researchers could benefit from running it on multiple machines.
Even if it isn’t obvious why a service suddenly became smarter and more accurate. But these improvements will eventually benefit ordinary people through Google’s everyday products. Because it’s only now we’ll start to really see what’s this unshackled TensorFlow is capable of.