keith2018/TinyTorch
A tiny deep learning training framework implemented from scratch in C++ that follows PyTorch's API.
TinyTorch helps deep learning engineers build and train neural networks from scratch, offering a highly optimized and low-level framework. It takes model architectures and training data, producing trained models for inference. This tool is for researchers and engineers who need fine-grained control over their deep learning stack or want to understand deep learning internals.
160 stars.
Use this if you need to develop highly optimized deep learning training routines in C++ with a familiar PyTorch-like interface, without external deep learning library dependencies.
Not ideal if you are looking for a high-level framework to quickly prototype deep learning models or do not require C++-level performance and control.
Stars
160
Forks
26
Language
C++
License
MIT
Category
Last pushed
Jan 29, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/keith2018/TinyTorch"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
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.