Tensors and Dynamic neural networks in Python with strong GPU acceleration

A Tensor module that allows a deep learning framework to switch seamlessly between different engines.

micro ml projects

Understanding neural network libraries and the automatic gradient computations (autograd) in the backward pass

A neural network library built around an Automatic Differentiation system written from scratch. The main focus of this project is its 'AutoGrad' system.

Learn the fundamentals of PyTorch in this hands-on lab from MIT’s Deep Learning course. You'll explore tensors, build neural networks, and implement gradient descent — all key building blocks for deep learning.

A simple Autograd engine written in Crystal

Re-implementation of micrograd by Andrej karpathy for learning purposes. Pytorch api like autograd and neural net library.

RUNE: RUsty Neural Engine

Solving Schrodinger's Equation with a Neural Network using numerical integration and autograd. Check https://arxiv.org/abs/2104.04795

Autograd minimal toy implementation in Python

Learn PyTorch basics with tensors, GPU checks, and tensor creation. Build a strong foundation for deep learning. 🌟 #GitHub

A minimal Java neural network library with autograd, MLP, and classification examples – built from scratch.

demo repository containing the experiments for my master's seminar @ TUM

Space-Time Attention with a Shifted Non-Local Search

Micrograd is an autograd built from scratch

A pedagogical implementation of Automatic Differation on multi-dimensional tensors.

Simple and basic , tiny autograd engine

A Small(Quant) Auto Grad Library with visualization of individual operations within Neurons