micrograd

Implementation of Karpathy's micrograd in Mojo 🔥

yet another scalar autograd engine - featuring complex numbers and fixed DAG

A lightweight autograd engine inspired by PyTorch and micrograd

🎓 super small and cute neural net engine, using only numpy

A simple vector valued autograd engine aimed to be between micrograd and tinygrad with a nn library.

An awesome collection about micrograd & friends - the tiny built from scratch (about 100 lines of code) autograd engine w/ a PyTorch-like neural net library on top

This is a minimal, from-scratch implementation of a scalar-valued automatic differentiation (autograd) engine and a small neural network library built on top of it. It follows the ideas behind Andrej Karpathy's micrograd, rebuilt line by line to actually understand how backpropagation works — not just call it.

A small, elegant and visulized implementation of reverse mode autograd that is distinct from Karpathy's micrograd. Zero dependency.

Nanograd is even smaller than micrograd | Autograd implementation with vanilla Python

A tiny scalar-valued autograd engine and a neural net library on top of it with PyTorch-like API

# The is a micrograd-mini. Inspired by Andrej Karpathy.

A minimal vectorized automatic differentiation engine for learning how backpropagation, broadcasting, and neural networks work under the hood.

This repository is a clean, educational reimplementation of Andrej Karpathy's micrograd project. It includes: - A minimal autograd engine (`Value` class) built from scratch - Neuron, Layer, and MLP abstractions built directly on top of `Value` - Manual training loop using scalar-based backpropagation along with Visualization

An unapologetic mock-off of the awesome micrograd.

A simple autograd engine built from scratch in Python that computes gradients automatically and powers neural network training using backpropagation.

A lean, but mean, autograd engine! (⚠️ not active anymore, since this was more of a educational project for me.)

numpy written in python from scratch | moved to delveopers/Axon

A tiny Python deep learning framework with autograd and MLP, built from scratch.

Implementation of Neural Network service with automatic gradient calculations

Micrograd is a small automatic differentiation engine and neural network library built with Python, designed for educational purposes. It demonstrates core concepts like automatic differentiation, backpropagation, and neural network building blocks, with testing against PyTorch.