TensorFlow 2.4 Installation on Nvidia Jetson Nano

Python API Installation

1. Update and Upgrade System Packages

sudo apt update
sudo apt upgrade

Ensures your system is up-to-date.

2. Install pip and pip3

sudo apt install python-pip python3-pip

Installs pip for Python 2 and Python 3.

3. Uninstall and Reinstall Numpy

sudo -H pip3 uninstall numpy && sudo -H pip3 install numpy==1.18.5

Ensures the correct version of Numpy is installed.

4. Remove Old TensorFlow Versions

sudo pip uninstall tensorflow && sudo pip3 uninstall tensorflow

Removes any existing TensorFlow installations.

5. Install Dependencies

sudo apt install gfortran -y &&\
sudo apt install libhdf5-dev libc-ares-dev libeigen3-dev -y &&\
sudo apt install libatlas-base-dev libopenblas-dev libblas-dev -y &&\
sudo apt install liblapack-dev -y &&\
sudo -H pip3 install Cython==0.29.21

Installs necessary dependencies.

6. Install H5py with Specific Cython Version

sudo -H pip3 install h5py==2.10.0
sudo -H pip3 install -U testresources numpy

Installs H5py with specific Cython version.

7. Upgrade Setuptools and Install Additional Packages

sudo -H pip3 install --upgrade setuptools
sudo -H pip3 install pybind11 protobuf google-pasta
sudo -H pip3 install -U six mock wheel requests gast
sudo -H pip3 install keras_applications --no-deps
sudo -H pip3 install keras_preprocessing --no-deps

Upgrades setuptools and installs additional packages.

8. Install gdown to Download from Google Drive

sudo -H pip3 install gdown

Installs gdown.

9. Download TensorFlow Wheel

gdown https://drive.google.com/uc?id=1DLk4Tjs8Mjg919NkDnYg02zEnbbCAzOz

Downloads TensorFlow wheel from Google Drive.

10. Install TensorFlow

sudo -H pip3 install tensorflow-2.4.1-cp36-cp36m-linux_aarch64.whl

Installs TensorFlow.

TensorFlow 2.4 C++ API Installation

1. Update and Upgrade System Packages

sudo apt update
sudo apt upgrade

Ensures your system is up-to-date.

2. Remove Old TensorFlow Versions

sudo rm -r /usr/local/lib/libtensorflow*
sudo rm -r /usr/local/include/tensorflow

Removes any existing TensorFlow installations.

3. Install Dependencies

sudo apt install wget curl libhdf5-dev libc-ares-dev libeigen3-dev
sudo apt install libatlas-base-dev zip unzip

Installs necessary dependencies.

4. Install gdown to Download from Google Drive

pip install gdown

Installs gdown.

5. Download TensorFlow C++ API Tarball

gdown https://drive.google.com/uc?id=1zJ_EF2aFkr8JU8JgTLfKMxC6KxE3DRD4

Downloads TensorFlow C++ API tarball from Google Drive.

6. Unpack the Tarball

sudo tar -C /usr/local -xzf libtensorflow-2.4.1-JetsonNano.tar.gz

Unpacks the tarball.

Now your Nvidia Jetson Nano is equipped with TensorFlow 2.4, ready to empower your machine learning projects!

Last modified March 18, 2024: omg (df686ea)