rwilliamspbg-ops/Sovereign_Map_Federated_Learning

Sovereign Map is a production-grade, Byzantine-tolerant Federated Learning framework. Utilizing the Mohawk Protocol for streaming aggregation, it achieves a 224x memory reduction, enabling secure orchestration of 100M+ nodes via TPM 2.0 hardware-rooted trust. Features full-stack observability with Prometheus & Grafana, built-in tokenomics telemetry

2 best points · 0 comments · 1 Hacker News post · 4 GitHub stars · View on GitHub

Hacker News discussions

• Show HN: I solo-validated Fed learning at 10M nodes with 50% Byzantine tolerance (2 points, 0 comments, by rwilliamspbgops, 2026-02-25)