learning-at-home/hivemind
A PyTorch library enabling decentralized deep learning training across geographically distributed volunteer computers.
Not currently ranked — collecting fresh signals.
star history
Hivemind enables training large neural networks across hundreds of computers from different organizations without a central coordinator. It uses a Distributed Hash Table for peer-to-peer communication, implements fault-tolerant backpropagation, and supports decentralized Mixture-of-Experts for scaling models beyond single machine memory. The library targets academic and research use cases where participants contribute computational resources.
Frequently asked
- What is learning-at-home/hivemind?
- A PyTorch library enabling decentralized deep learning training across geographically distributed volunteer computers.
- Is hivemind open source?
- Yes — learning-at-home/hivemind is open source, released under the MIT license.
- What language is hivemind written in?
- learning-at-home/hivemind is primarily written in Python.
- How popular is hivemind?
- learning-at-home/hivemind has 2.5k stars on GitHub.
- Where can I find hivemind?
- learning-at-home/hivemind is on GitHub at https://github.com/learning-at-home/hivemind.