A self-contained course to learn the basics of neural networks: from backprop to GPT!
Visit micrograduate-book.github.io.
| # | 📔 | 🔗 |
|---|---|---|
| 1 | micrograd |  |
| 2 | makemore1 | |
| 3 | makemore2 | |
| 4 | makemore3 | |
| 5 | makemore4 | |
| 6 | makemore5 | |
| 7 | picoGPT | |
- Clone repo:
git clone https://github.com/ckaraneen/micrograduate.git- Create and activate env:
cd micrograduate
conda env create --file environment.yaml
conda activate micrograduate-env- Install requirements:
uv pip install -r requirements.txt- Run notebooks in Notebook Server, VSCode/Cursor, etc.
microgra∇uate builds on the works of Andrej Karpathy (check out Eureka Labs!) and Jay Mody:
- Neural Networks: Zero to Hero lecture series and repo
- micrograd
- makemore
- An Intuition for Attention
- GPT in 60 lines of code
- picoGPT
It would not exist without their contributions. ❤️