View Code? Open in Web Editor
NEW
Implementation of asynchronous federated learning in flower.
License: Apache License 2.0
flower-async's Introduction
๐ Hi, Iโm @r-gg
๐ Iโm interested in Software Engineering, Distributed / Cloud Computing and Machine Learning
๐ฑ Iโm currently studying at TU Wien (Master: Software Engineering and Internet Computing)
flower-async's People
Contributors
Watchers
flower-async's Issues
Can you provide an example or explanation about AsynchronousStrategy.
Thanks!!
I am trying to get flower-async running in the simulation mode. After a bit of tinkering, I am stuck with the following error.
INFO : Sampling failed: number of available free clients (1) is less than number of requested free clients (2).
log (INFO , "Initializing global parameters" )
self .parameters = self ._get_initial_parameters (timeout = timeout )
log (INFO , "Evaluating initial parameters" )
res = self .strategy .evaluate (0 , parameters = self .parameters )
As I understand it, it is due to the client not getting freed after being sampled for the initial parameters.
Any insight into how to fix the same / Have you gotten the simulation running?
Do I need to write a custom simulation class to get it working?