ABOUT
Bytes and Brains is building a substrate for networked machine learning. A framework that takes an authored workload, partitions it across nodes, and ships the bytes between them. The goal is one substrate where every distributed ML strategy (federated, gossip, peer-to-peer, split) composes as a binding on the same surface.
The library is on crates.io. The substrate design is documented in the whitepaper and the source is on GitHub. The work is happening in public. Researchers, practitioners, and operators are invited to read along and reach out.
Machine learning that runs outside the data center, where the data already lives. Workloads authored once and partitioned across whatever network the operator can reach: phones, sensors, regulated environments, on-prem fleets, peer-to-peer overlays. The framework owns the bytes between nodes. The host owns its sockets. The field's strategies compose on the same foundation rather than fragmenting into one library per paradigm.
The whitepaper is the idea, the design, and the direction. The research page lists the papers behind the substrate and the protocols built on it. The blog carries deeper writing on the algorithms and primitives. Source is on GitHub.
Founded by Matthew Love.
Contact:
[email protected]
The library is open source under AGPL-3.0. A commercial license is available for production deployments that need to be exempt from the AGPL's source-disclosure obligations. See the license page for terms, or email about your deployment.