cdgriffith/PyCon-AI-Crash-Course

Course materials

PyCon US 2025 Tutorial - AI Crash Course for Python developers

This repo contains the exercises for Anthony Shaw's PyCon US Tutorial AI Crash Course for Python Devs

Setup

You will need Jupyter to run these exercises. I recommend installing the Jupyter Extension in VS Code if you don't already have a Jupyter environment

After that, you'll need to create a virtual environment and install the packages in requirements.txt:

$ pip install -r requirements.txt

Lessons

These are the three primary lessons we will be working through:

• Lesson 1: LLMs, tokenizers, and prompts
• Lesson 2: Embedding models
• Lesson 3: Function, tools, and agents

Customizing Game Play

This repo contains multiple references to a made-up game. If you want to customize that game with your own, you'll need to:

1. Modify utils/game.py and change RULES
2. Modify utils/game.py and update determine_winner_programmatically() with your new logic
3. Run make_training_data_csv() in utils/game.py to make a new utils/game_training.csv file

Note this isn't required, but you want to do this yourself later on.