EPFL Logo CENTER FOR
DIGITAL TRUST
Garfield

Garfield

System support for byzantine machine learning

Garfield is a library to build Byzantine machine learning (ML) applications on top of popular frameworks such as TensorFlow and PyTorch. We show how to use Garfield to build different architectures for ML applications like single server, multiple workers (SSMW), multiple servers, multiple workers (MSMW), and fully decentralized architecture.

Byzantine ResilienceDecentralizedDistributed LearningFederated LearningPyTorchTensorFlow
Key facts
Maturity
PrototypeIntermediateMature
Support
C4DT
Active
Lab
Unknown
  • C4DT work
  • Technical
  • Research papers
  • Presentation
  • Demo
Status: Active
Timeline: Started in Spring 2021, demo in Autumn.

Distributed Computing Lab

Distributed Computing Lab
Rachid Guerraoui

Prof. Rachid Guerraoui

The Distributed Computing Lab focuses currently on Scalable Implementations of Cryptocurrencies, Byzantine fault tolerance and privacy in distributed machine learning, distributed algorithms making use of RDMA and NVRAM.
Go back

This page was last edited on 2024-03-22.