Jetson Generative AI Lab

Document generation status:

About this repo

This repo is to host a tutorial documentation site for running generative AI models on NVIDIA Jetson devices.

The auto generated documentation is hosted on the following, using their CI/CD feature to automatically generate/update the HTML documentation site upon new commit:

• GitHub Pages site

How to use this repo locally

Option 1: Docker Setup (Linux systems with Docker)

Initial setup

https://squidfunk.github.io/mkdocs-material/getting-started/

sudo apt install -y docker.io
sudo docker pull squidfunk/mkdocs-material

Start development server on http://localhost:8000

docker run --rm -it -p 8000:8000 -v ${PWD}:/docs squidfunk/mkdocs-material

Option 2: Native Python Setup (Recommended for Windows)

Initial setup

1. Install Python 3.8+ and pip (if not already installed):

   • Windows: Download from python.org or use Microsoft Store

2. Install MkDocs Material and dependencies:

   pip install mkdocs-material
   pip install mkdocs-redirects
   pip install beautifulsoup4
   pip install lxml

Start development server on http://localhost:8000

mkdocs serve

Build the site

mkdocs build

Test the post-processing

docker run --rm -it -p 8000:8000 -v ${PWD}:/docs squidfunk/mkdocs-material build
pip install beautifulsoup4
pip install lxml
python3 ./scripts/duplicate_site_with_postprocess.py ./site ./site_postprocessed
sudo apt install python3-livereload
livereload ./site_postprocessed

Troubleshooting

Native Python Setup Issues

Windows:

• If mkdocs command is not found, ensure Python Scripts directory is in your PATH
• Use python -m mkdocs serve if the direct command doesn't work
• For corporate networks, you may need to use pip with proxy: pip install --proxy http://proxy:port package-name

Jetson:

• If you get permission errors, use pip install --user to install packages locally
• Ensure you have sufficient disk space for the Python packages

Docker Setup Issues

If you get "docker: Got permission denied while trying to connect to the Docker daemon socket at ..." error, issue sudo usermod -aG docker $USER; newgrp docker to get around with the issue.