Repolode
Submit
tensorflow

tensorflow/tensorflow

An end-to-end open source platform for building and deploying machine learning models, with Python and C++ APIs.

View on GitHubWebsite
Stars
201,197
Forks
79,842
Open issues
3,441
Last commit
4 hours ago

What is it?

TensorFlow is an end-to-end open source platform for machine learning, originally developed by researchers and engineers on Google's Machine Intelligence team. It covers the full workflow from model research and training to deployment, with a surrounding set of tools, libraries and extensions.

The project offers stable Python and C++ APIs, and a less strictly compatible API for other languages. Installation is available as a pip package, a CPU-only variant, Docker containers, or a build from source, with GPU support for CUDA-enabled NVIDIA cards on Ubuntu and Windows and device plugins for other hardware such as DirectX and Apple Metal.

TensorFlow is used for neural network research and for building ML-powered applications, and its ecosystem includes TensorBoard for visualization and TensorFlow Lite (LiteRT) for mobile and Android deployment.

Key features

  • 01Stable Python and C++ APIs for building and training models
  • 02GPU acceleration on CUDA-enabled NVIDIA cards, plus device plugins for other hardware
  • 03Installable via pip, Docker, or from source, with CPU-only builds available
  • 04Ecosystem tools including TensorBoard for visualization and LiteRT for mobile deployment

Use cases

  • Training and researching neural network models
  • Building ML-powered applications that need to be deployed to production
  • Running inference or training on Android and other mobile devices
  • Visualizing training metrics and model graphs with TensorBoard

Quick start

  1. 1

    Install the TensorFlow pip package

    pip install tensorflow
  2. 2

    Start Python in your terminal

    python
  3. 3

    Run a first TensorFlow program that adds two numbers

    import tensorflow as tf tf.add(1, 2).numpy()

Things to know

  • GPU support is limited to Ubuntu and Windows for CUDA cards; other devices need separate device plugins.
  • The README describes a non-guaranteed backward compatible API for languages other than Python and C++.

Topics

#deep-learning#machine-learning#neural-network#python#cpp#gpu#mobile#deep-neural-networks#distributed#ml

FAQ

What is tensorflow?
An end-to-end open source platform for building and deploying machine learning models, with Python and C++ APIs. TensorFlow is an end-to-end open source platform for machine learning, originally developed by researchers and engineers on Google's Machine Intelligence team. It covers the full workflow from model research and training to deployment, with a surrounding set of tools, libraries and extensions.
Is tensorflow free to use? What license does it have?
tensorflow is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is tensorflow still maintained?
Actively maintained. The last commit to tensorflow was on Oct 20, 2026.
What can I use tensorflow for?
Training and researching neural network models Building ML-powered applications that need to be deployed to production Running inference or training on Android and other mobile devices Visualizing training metrics and model graphs with TensorBoard
How do I install tensorflow?
Install the TensorFlow pip package: pip install tensorflow Start Python in your terminal: python Run a first TensorFlow program that adds two numbers: import tensorflow as tf tf.add(1, 2).numpy()
Where can I get tensorflow?
The source code and installation instructions are on GitHub: https://github.com/tensorflow/tensorflow

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects