Learning Tinygrad

work in progress 🚧

First I had to become a user. I followed Andrej Karpathy’s Makemore series using Tinygrad. Below, are links to models I created along the way.

Lesson Tinygrad Implementation Notes
Makemore: Bigram model ✅ Count Model, ✅ NN Model no load/store
Makemore: MLP ✅ MLP Model
Makemore: BatchNorm ✅ BatchNorm Model
Makemore: Backprop Ninja N/A
Makemore: WaveNet ✅ WaveNet Model

Next, I needed to learn how to build the project. It’s pretty straightforward.

git clone https://github.com/tinygrad/tinygrad.git
cd tinygrad
python3 -m venv venv && source venv/bin/activate 
python3 -m pip install -e .

Now to supplement with media. I’m creating Anki flashcards from official and semi-official docs to internalize core concepts.


TODO… reimplementing recent commits.

-Mitchell Hashimoto
Commit Reimplementation
TODO TODO

TODO: Make a small, bite size change.