What is TensorFlow?

TensorFlow is an open-source software suite to design and build deep learning models. It is a software library for doing complex numerical computations quickly using data-flow graphs. It was originally developed by the Google Brain team for internal deep learning and neural networks research. In November 2015, it was released under the Apache 2.0 open-source license. TensorFlow supports all major operating systems like Linux, macOS, Windows, alongside iOS and Android. It has a highly flexible architecture that enables easier deployment of computation models from desktops, mobiles, and servers to different types of platforms, such as CPUs, GPUs and TPUs (Tensor Processing Units).

Points to Remember

  • TensorFlow expresses its complex numerical computations as data-flow graphs that can be applied to several areas where deep learning is used. Image recognition, language processing, machine translation, production prediction, word embeddings are few examples of its applications.
  • TensorFlow is the most widely used framework for design, build, and improve deep learning models. It is used in many renowned enterprises like Google, Airbus, IBM, Twitter and others.