kann
A lightweight C library for building and training small to medium artificial neural networks with minimal dependencies.
What is kann?
KANN is a lightweight C library for building and training small to medium artificial neural networks, including MLPs, CNNs, and RNNs. It implements graph-based automatic differentiation to create flexible models with features like weight sharing and multiple outputs. The library solves the need for a portable, dependency-free neural network toolkit in C/C++ environments.
Developers working in C/C++ who need to experiment with or deploy small to medium neural networks without external dependencies, or those wanting to learn deep learning library internals.
KANN offers a unique combination of flexibility and minimalism—it supports complex neural architectures while being under 4000 lines of code with no non-standard dependencies. It's faster than other lightweight frameworks and is ideal for CPU-only scenarios where portability and simplicity are critical.
Overview
A lightweight C library for artificial neural networks
Use Cases
Best For
- Experimenting with neural networks in pure C/C++
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
