Welcome
Welcome to the Advanced AI ModIA course page!
This course will give students the basis to understand modern AI systems, with a focus on implementation and applications. The course is structured around:
- Lectures
- Practical programming sessions
- One project that will start mid-semester.
In this course, you will learn about modern LLMs, VLMs, Self-Supervised Learning, and Reinforcement Learning involved in LLM training.
This page is the main hub for the course. You will find lecture notes, practical programming and project guidelines, and related code. This is the first year this course is given, so expect some changes and updates along the way. See below for how to update your local repository with the latest changes.
Repository Setup¶
First fork this repository to your own GitHub account, then clone it locally
git clone your_forked_URL
Then add this repo as a remote to pull updates
git remote add upstream https://github.com/paulnovello/Advanced-AI
To update your forked repo with the latest changes from this original repo, run:
git fetch upstream
git merge upstream/main
If you only make changes to the files specified in the practical programming sessions, you should never have merge conflict to manaually resolve. If you do, just reach out to the course staff.
Environment Setup¶
1. Install uv¶
We recomend to set up the environment once and for all with uv. Uv is a fast Python package installer and resolver. Install it using one of the following methods:
macOS/Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh
Windows:
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
Using pip:
pip install uv
2. Create env¶
uv env ./aai_env
3. Install Dependencies¶
Install the required dependencies using uv:
uv sync --group student
4. Activate env¶
Activate the virtual environment created by uv:
source ./aai_env/bin/activate # macOS/Linux
.\aai_env\Scripts\activate # Windows
or use uv's run command to execute scripts within the environment without activating it:
uv run python # my_script.py