fedlib-beta1 – A Federated Learning Experimental Python Library

fedlib – A Federated Learning Experimental Python Library (Beta 1)

Running federated learning (FL) algorithms requires considerable efforts to set up the decentralized running environments in terms of both hardware and software dependencies. It raises great challenges for Researchers to test their ideas. Especially in scaled federated learning, where thousands of edge clients may involved into communication, it would cause researcher spend tons of energy to configure every single edge client.

In fedlib, we aim to provide federated learning engineers and developers a user-friendly federated learning platform to quickly test and run their customized FL algorithm. Additionally, we also target to provide developers easy to use APIs for deploying specialized FL models to real-world applications.