lucasdelimanogueira/PyNorch
Recreating PyTorch from scratch (C/C++, CUDA, NCCL and Python, with multi-GPU support and automatic differentiation!)
This project offers a foundational deep learning framework, designed to provide insights into the inner workings of such systems. It takes raw numerical data, typically in tensor format, and allows you to build, train, and run neural network models, outputting trained models and gradient information. It's intended for machine learning engineers and researchers who want to understand the underlying mechanics of deep learning frameworks.
164 stars.
Use this if you are a deep learning practitioner or student eager to explore and learn how frameworks like PyTorch handle tensors, automatic differentiation, and multi-GPU training from scratch.
Not ideal if you need a production-ready deep learning framework with a comprehensive set of features, optimized performance, and extensive community support for real-world applications.
Stars
164
Forks
11
Language
Python
License
—
Category
Last pushed
Nov 25, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/lucasdelimanogueira/PyNorch"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
apache/tvm
Open Machine Learning Compiler Framework
uxlfoundation/oneDNN
oneAPI Deep Neural Network Library (oneDNN)
Tencent/ncnn
ncnn is a high-performance neural network inference framework optimized for the mobile platform
OpenMined/TenSEAL
A library for doing homomorphic encryption operations on tensors
iree-org/iree-turbine
IREE's PyTorch Frontend, based on Torch Dynamo.