Search Engine and Technology giant Google on Thursday announced the release of its most advanced machine learning library TensorFlow 2.0 Alpha version. At the 3rd annual TensorFlow Developer Summit; Google announced TensorFlow 2.0 Alpha, Coral Development Platform, Coral Dev Board and USB accelerator. These products will further fuel the growth of TensorFlow machine learning library.
According to the Google TensorFlow 2.0 Alpha version is simple, intuitive and feature rich framework for developing and deploying large scale Deep Learning models. The Alpha Version of TensorFlow 2.0 introduced new privacy focused tool for privacy aware model training in TensorFlow. New privacy focused tools added to TensorFlow 2.0 Alpha includes TF Federation and TF Privacy. This will enable developer train model while considering user’s data privacy into account.
TensorFlow is very popular machine learning and deep learning framework from Google; this framework is widely used and till now it recorded 41 million times download. Google already announced the top features of TensorFlow 2.0 and now alpha version of this library is available for developers for testing.
Google also announced TensorFlow for JavaScript developers and it is available with the name TensorFlow.js which can run in web browser. This version of TensorFlow brings machine learning to the web browser where JavaScript developers can use full potential of machine learning to develop browser based applications.
The privacy-tools introduced in TensorFlow 2.0 alpha are TF Federation and TF Privacy. These tools will help developer to train model with decentralised data set without worrying about user privacy. These tools enable developer to train model in fairer and privacy safe manner.
TensorFlow 2.0 provide support for better integration with the tf.keras library which is higher level API for developing machine learning and deep learning models. This will help developers in quickly prototype and build machine learning models. After completion of data ingestion and cleaning developers can use keras library with TensorFlow for building application quickly. Google also launched TensorFlow Datasets which includes many common datasets frequently used in machine learning.
TensorFlow 2.0 top features includes easy model building, better integration with Keras library, robust model deployment in production across platform, powerful features for research work, cleaning of deprecated API and support for privacy focus model training. TensorFlow is one of the most used machine learning, deep learning and artificial intelligence library available for data scientists. Google recently released “Coral Dev Board” and USB accelerator with in-built TPU (Tensor Processing Unit) which can be used for developing Edge computing infrastructure. TensorFlow light is the programming library which can be used for application development. With the release Google is expending the market of TensorFlow to Edge computing also.
12,903 total views, 11 views today
