Privacy-Preserving
Federated Learning
syft-flwr is an open-source framework that combines Flower's federated learning capabilities with file-based communication (e.g. Google Drive, SyftBox). Train machine learning models collaboratively across distributed datasets without centralizing data—with easy setup, offline capability, and no servers required.
Simplified FL Development
Build federated learning applications with minimal code changes to existing Flower projects—just bootstrap and deploy.
File-Based Communication
Train models collaboratively without requiring direct network connections—communication happens via file sync (Google Drive or SyftBox), enabling offline-capable workflows.
Zero Infrastructure
No servers to maintain, no complex networking setup—just notebooks and file sync (Google Drive or SyftBox). Get started in minutes.
Privacy by Design
Data never leaves its source—only model updates are shared, with full consent management and privacy-preserving aggregation.
Train models across distributed datasites without centralized servers—using file-based sync for asynchronous message passing between participants.
Built on Flower's robust FL framework—supports FedAvg, custom strategies, and all standard Flower features with minimal code changes.
Run local simulations for development and testing, then seamlessly deploy to distributed production environments.
Raw data never leaves the data owner's environment—only model parameters are exchanged, with explicit consent required for each training job.
federated learning with privacy?