Mentions légales du service

Skip to content

Fedprox included in torchnn

BALELLI Irene requested to merge feature/FedProx_training_plan into develop

FedProx optimization scheme is included in torch training plan and can be enabled by passing the FedProx_mu parameter in model_args, as shown in the notebook pytorch-MNIST-FedProx.ipynb with the basic MNIST example (otherwise standard optimization scheme will be performed).

Merge request reports