SawyerAlston/MNIST-NN-Pure-Math

A "from-scratch" 2-layer neural network for MNIST classification built in pure NumPy, featuring mini-batch gradient descent, momentum, L2 regularization, and evaluation tools — no ML libraries used.

14
/ 100
Experimental

No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 3 / 25
Maturity 9 / 25
Community 0 / 25

How are scores calculated?

Stars

3

Forks

—

Language

Python

License

MIT

Category

numpy-based-neural-networks

Last pushed

Jul 09, 2025

Commits (30d)

0

GitHub
NumPy-Based Neural Networks · 124 frameworks

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/SawyerAlston/MNIST-NN-Pure-Math"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.

Higher-rated alternatives

digantamisra98/Mish

Official Repository for "Mish: A Self Regularized Non-Monotonic Neural Activation Function" [BMVC 2020]

66

Sentdex/nnfs_book

Sample code from the Neural Networks from Scratch book.

51

itdxer/neupy

NeuPy is a Tensorflow based python library for prototyping and building neural networks

51

vzhou842/cnn-from-scratch

A Convolutional Neural Network implemented from scratch (using only numpy) in Python.

50

nicklashansen/rnn_lstm_from_scratch

How to build RNNs and LSTMs from scratch with NumPy.

49

Explore ML Frameworks

All categories Trending ML Framework directory Insights