RobustFL – Simulate Byzantine Attacks in Federated Learning

An open research tool for testing and benchmarking poisoning attacks in FL — without complex setup or infrastructure.

• ✅ Built for researchers, educators, and students
• ✅ Supports common attacks: label flipping, model replacement, Sybil
• ✅ No server or GPU setup needed
• ✅ REST API for automated testing
• ✅ Open-source, extensible backend