Skip to content

sujeendran/tensorflow-addons

 
 

Repository files navigation




PyPI Status Badge PyPI - Python Version Documentation Gitter chat Code style: black

Fork made for NVIDIA Jetson Nano Jetpack 4.4

This fork has changes made to build tensorflow-addons for Tensorflow 2.3 Jetson Nano built from source. Should work with >=TF2.2. Does not work with NVIDIA provided Tensorflow 2.1 build since a lot of API calls seems to be changed.

For detailed documentation refer to original repo: https://github.com/tensorflow/addons

Quick Start

git clone https://github.com/sujeendran/tensorflow-addons.git
cd addons

pip install artifacts/tensorflow_addons-*.whl

Installing from Source

You can also install from source. This requires the Bazel build system (version >= 1.0.0).

If you have jetson_stats installed, you can use sudo jtop and enable the extra swap space as the bazel build can completely hang your system for a while(but should still build in the background very slowly even if you dont enable swap)

GPU and CPU Custom Ops
git clone https://github.com/sujeendran/tensorflow-addons.git
cd addons

export TF_NEED_CUDA="1"

# Set these if the below defaults are different on your system
export TF_CUDA_VERSION="10.2"
export TF_CUDNN_VERSION="8"
export CUDA_TOOLKIT_PATH="/usr/local/cuda"
export CUDNN_INSTALL_PATH="/usr/lib/aarch64-linux-gnu"

# This script links project with TensorFlow dependency
python3 ./configure.py

bazel build --enable_runfiles build_pip_pkg
bazel-bin/build_pip_pkg artifacts

pip install artifacts/tensorflow_addons-*.whl

About

Useful extra functionality for TensorFlow 2.x maintained by SIG-addons

Resources

License

Code of conduct

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 77.6%
  • C++ 12.2%
  • Smarty 5.1%
  • Starlark 4.0%
  • Other 1.1%