Backpropagation Python Packages

autograd

Efficiently computes derivatives of NumPy code.

hypercube-lcn

Locally connected deep neural network on a Boolean hypercube: the cube's edges are the wiring, and every weight trains. C++ core + Python SDK.

termux-train

📱 Ultra-lightweight On-Device Tensor & DAG Autograd Deep Learning Framework for Android ARM64 Termux. SafeTensors zero-copy, LoRA fine-tuning, and RoPE Transformers.

neural-network-digits

Neural network from scratch in pure Python + NumPy that recognises handwritten digits (MNIST), with an educational GUI: watch it learn, tweak it while it trains, look inside every layer.

odtbrain

Python library for diffraction tomography with the Born and Rytov approximations

pyrddlgym-jax

JAX compilation of RDDL description files, and a differentiable planner in JAX.

weed-loader

Minimalist AI/ML inference and backprogation in the style of Qrack

weed-loader-cpu

Minimalist AI/ML inference and backprogation in the style of Qrack

ntk-ml

A from-scratch neural network and tensor library built in Python for learning, experimentation, and deep understanding of modern ML systems.

synapx

A deep learning library powered by a C++ autograd engine, featuring a PyTorch-like API and cross-platform support.

synapgrad

Lightweight autograd engine with a deep learning library and a Torch-like API, built from scratch.

npnet

Build neural networks based only on Numpy

pileoffeather

Lightweight and easy to use ml library for small projects, create a neural network in minutes.

mangograd

🥭 A PyTorch-shaped, NumPy-powered autograd engine you can actually read. Born from watching Karpathy's micrograd while eating a mango.

dlfs

Implementation from scratch (using numpy arrays) of a framework based on keras interface which allows to build and train Fully Connected Networks and Convolutional Neural Networks (CNNs).

freegrad

PyTorch extension for alternative backward rules and gradient transforms (STE, gradient jamming, non-standard activations).

tonygrad

A tiny scalar-valued autograd engine with a small PyTorch-like neural network library on top.

symple

Automatic differentiation in python

autograd-latest

Efficiently computes derivatives of NumPy code.

nanograd-lite

Automatic differentiation engine built from scratch using only numpy with computational graphs and backpropagation.

minitf

Simplified version of Tensorflow for learning purposes.

randomized-based-feedforward-neural-network

Pytorch implementation of a Randomized-Based FFNN

pocketgrad

A lightweight autograd engine and neural network library.

sympyle

Automatic differentiation in python