Autograd Python Packages

torch

Tensors and Dynamic neural networks in Python with strong GPU acceleration

autograd

Efficiently computes derivatives of NumPy code.

autograd-gamma

NotImplementedError: VJP of gammainc wrt argnum 0 not defined

pennylane

PennyLane is an open-source quantum software platform for quantum computing, quantum machine learning, and quantum chemistry. Create meaningful quantum algorithms, from inspiration to implementation.

torch-sparse

PyTorch Extension Library of Optimized Autograd Sparse Matrix Operations

essos

e-Stellarator Simulation and Optimization Suite

lucid-dl

Lumerico's Comprehensive Interface for Deep Learning

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.

neuralengine

An open-source framework/library for building and training neural networks in Python with support for GPUs (CUDA).

legume-gme

🌱 Guided-mode expansion of photonic crystal slabs

nutens

Tensor based engine for calculating neutrino oscillation probabilities in a fast, flexible, and differentiable way

pyborch

PyTorch's API that trains in a browser tab on WebGPU — autograd, nn, optim, capture/replay, int8, ONNX out; held to real torch's values on 4,764 golden cases; measured on one page against TF.js, jax-js, Burn and ONNX Runtime Web (4–12× faster training, ahead of ORT at inference)

rustforge-rl

A pure Rust Reinforcement Learning framework featuring a custom tensor engine, autograd system, and modular RL algorithms.

memzero-train

Zero-memory, multi-core-CPU local AI training for ordinary laptops.

pyerrors

Error propagation and statistical analysis for Monte Carlo simulations in lattice QCD and statistical mechanics using autograd.

norse

Deep learning with spiking neural networks (SNNs) in PyTorch.

torchtomo

Differentiable 2D CT in PyTorch with parallel/fan-beam projectors, exact adjoints, filtered backprojection, learnable scan geometry, and optional CUDA acceleration.

neuralfabric

The Machinery Behind Machine Learning.

mygrad

Drop-in autodiff for NumPy.

boost-loss

Utilities for easy use of custom losses in CatBoost, LightGBM, XGBoost.

nabla-ml

Nabla: High-Performance Scientific Computing

pyember

ML Library from scratch with only standard libraries

agent-torch

large population models

nets

Recreating PyTorch from scratch, using Numpy. Supports FCN, CNN, RNN layers.