shainet
A pure Crystal machine learning library for building and training neural networks with CPU/GPU support and PyTorch compatibility.
What is shainet?
SHAInet is a neural network library written in pure Crystal, designed for building and training machine learning models. It provides tools for creating various network architectures, supports both CPU and GPU computation, and includes features for handling large datasets and importing models from frameworks like PyTorch. The library aims to simplify neural network development while maintaining performance and flexibility.
Crystal developers and researchers interested in machine learning, especially those looking for a native Crystal solution for neural networks or needing GPU-accelerated model training within the Crystal ecosystem.
Developers choose SHAInet for its seamless integration with Crystal, offering a pure-Crystal implementation with optional GPU acceleration via CUDA. Its ability to import PyTorch and HuggingFace models provides interoperability, while features like streaming data support make it practical for real-world, large-scale applications.
Overview
SHAInet - a pure Crystal machine learning library
Use Cases
Best For
- Building neural networks entirely within the Crystal programming language
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.
