To learn the deep learning algorithms, I changed my home computer into dual boot machine with windows 10 and pop! os 22.04

Install cuda tools

Make an account in nvidia home page and you can download cuda tools. As of today (22/07/14), I can download cuda 11.7 version. If you choose options like linux, x86_64, Ubuntu, 22.04, deb(network), then you will get the following commands instructions.

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin
sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/3bf863cc.pub
sudo add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/ /"
sudo apt-get update
sudo apt-get -y install cuda

Then you can install cuda tools.

Install cuDNN

For deeplearning pipeline, you also need to install cuDNN and you can find the link here Please log in to nvidia website first. As of today (22/07/14), you can download two versions - cuDNN v8.4.1 for CUDA 11.x and cuDNN v8.4.1 for CUDA 10.2

you can find the details of the instructions here

sudo dpkg -i cudnn-local-repo-${OS}-8.x.x.x_1.0-1_amd64.deb
sudo cp /var/cudnn-local-repo-*/cudnn-local-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get install libcudnn8=8.x.x.x-1+cudaX.Y
sudo apt-get install libcudnn8-dev=8.x.x.x-1+cudaX.Y
sudo apt-get install libcudnn8-samples=8.x.x.x-1+cudaX.Y

You can also reference the following site

Compile tensorflow

Once you install cuda and cudnn, you can start to compile tensorflow. Before you start, please follow instructions in here The key component is to install go and bazelisk.

Compile Jax

You can also compile jax using this site. Please make sure to install jaxlib and jax.