Monday, June 03, 2019

HOWTO : Install Keras On Nvidia Jetson Nano Developer Kit

To install JetPack 4.2 on Nvidia Jetson Nano Developer Kit, you need to follow this link.

Since JetPack 4.2 is using Ubuntu 18.04 instead of Ubuntu 18.04.2, Unity is installed by default. I prefer to uninstall Unity and get back the GNOME 3.

sudo apt update
sudo apt -y dist-upgrade
sudo apt remove unity-session unity
sudo apt install -y ubuntu-session gdm3 firefox gparted chrome-gnome-shell gnome-tweak-tool nano
sudo apt -y autoclean
sudo apt -y autoremove


Reboot the box.

If you have ownCloud :

sudo apt install -y owncloud-client

** owncloud requires you to enter password every time when boot.

To install Gnome Shell Extensions :

Harddisk LED to display the activity of the hard drive/SSD. It is recommended for this developer kit.

To set "Problem Reporting" to "Automatic" at "Privacy" of "Settings" in order to prevent unexpected popup windows.

To install Keras :

sudo apt-get install libhdf5-serial-dev hdf5-tools
sudo apt install -y python3-pip python3-dev python3-scipy
sudo apt-get install zlib1g-dev zip libjpeg8-dev libhdf5-dev

sudo pip3 install -U pip
sudo pip3 install -U numpy grpcio absl-py py-cpuinfo psutil portpicker six mock requests gast h5py astor termcolor

sudo pip3 install --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v42 tensorflow-gpu

sudo pip3 install -U keras


To test if it works or not :

python3

>>> import keras

If there is no error message and showing "Using TensorFlow backend.", it works. To quit it :

>>> quit()

If you want to create swap file, you may need to use this resources.

That's all! See you.