
MacOS 10.12.6 (Sierra) or later 64-bit version with no GPU support. Pip 19.0 or later with manylinux2010 support for Linux users. The 64-bit version of Ubuntu 16.04 or later. This needs to come with Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019 The 64-bit Windows 7 or a later version. Python 3.7-3.10 (Python 3.6 used to be included, but now the requirements start with version 3.7). Verify that your system meets the following requirements. If you want to install Tensorflow (TF) to access this library from Google AI, we will walk you through what you need to know about getting your machines set up properly.įor you to install Tensorflow on your system, you will need to make sure you meet the minimum system requirements, including the required version of Python. The Python API then allows programmers to access all the information contained within Tensorflow.
The connections between the different processing nodes of the graph is known as a tensor or a multidimensional data array. Developers can then create dataflow graphs showing how the data collected will move through a graph. It combines neural networks, or bundles of machine learning and deep learning models.
Tensorflow was initially released in 2015, with an updated version available in 2019.
Tensorflow offers an open-source library to help developers create the deep-learning models and machine-learning frameworks to accomplish goals like sentiment analysis or image recognition. Machine learning has become an increasingly important part of how businesses and customers interact with the digital world.