autograd-engine
A reverse-mode automatic differentiation engine built from scratch in Python — no PyTorch, no TensorFlow. 97.63% on MNIST, 737 tests, 183 gradient checks, and 21 chapters explaining every derivative.
- autograd
- automatic-differentiation
- backpropagation
- deep-learning
- educational
- from-scratch
- gradient-descent
- machine-learning
- micrograd
- mnist
- neural-network
- numpy
- python
- Stars
- 6
- Forks
- 5
- + today
- +1
- Created
- 2mo
Ranking data as of November 2, 2026 (UTC).
Star History
Today, hour by hour
01
Overview
A reverse-mode automatic differentiation engine built from scratch in Python — no PyTorch, no TensorFlow. 97.63% on MNIST, 737 tests, 183 gradient checks, and 21 chapters explaining every derivative. It ranks #761 on GitTiger, gaining +1 star on November 2, 2026 (UTC).
The project is written in Python and has 5 forks. It was created 2mo ago.
Installation
git clone https://github.com/Adebayodamilola20/autograd-engine.git
cd autograd-engine
# see README for setup