Skip to content

ZK-Plus/dfl

 
 

Repository files navigation

Proof of concept implementation: Practical Verifiable Decentralized Federated Learning with TEEs

Advancing the Efficiency of Decentralized Federated Learning

The folders represent different components of the MVP.

  • Docker setup: Recommended.
  • Neural Network: ANN with additional FedAvg algorithm.
  • Node Server: Logic for the components inside the edge devices, encapsulates the Nural Network.
  • Remote attestation Code: zkVM-based remote attestation and Foundry Project including all smart contracts and interfaces.

Decentralized Federated Learning

Docker (Recommended)

We recommend you use the docker files. All other guides (non docker) are for leaning and documentation purpous.

See Docker README.

Other Readme´s

Importent configuration files

  • The .env file you have to create yourself has the round configuration (i.e. epoch, round)
  • The starter_docker file is a script that deployes the smart contracts and sets the initial values (i.e. register workers etc.)

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 34.5%
  • Makefile 18.7%
  • Solidity 16.3%
  • Rust 9.4%
  • JavaScript 8.1%
  • Fortran 5.1%
  • Other 7.9%