Interactive worksheets to practice the concepts in Andrej Karpathy's Neural Networks: Zero to Hero video series.
I recommend starting with the versions without solutions, and only consulting the provided solution for a given cell once 1) you've completed it or 2) you're stuck despite reviewing the relevant section of video.
IMPORTANT NOTE: Make sure to save a copy of the worksheet (File -> Save a copy in Drive) before starting work on it, and verify that you can find the copy you've made.
| Worksheet | Colab | .py | Colab (w/solutions) | .py (w/solutions) |
|---|---|---|---|---|
| 02a-building_makemore_probability_table |  |
link | |
link |
| 02b-building_makemore_slp | |
link | |
link |
| 03-building_makemore_mlp | |
link | |
link |
git clone [email protected]:Russ741/karpathy-nn-z2h.git
cd karpathy-nn-z2h
conda env create -f environment.yml
conda activate karpathy-nn-z2h
pre-commit install
jupyter notebookThen open Jupyter in your browser using the link it printed out in the console, and navigate to the desired worksheet.
To update the conda environment when the environment.yml file is updated:
conda env update -f environment.yml