ResearchBenchmarksImage Classification on CIFAR-100 2009 (test) (Accuracy and Efficiency)Image Classification on CIFAR-100 2009 (test) (Accuracy and Efficiency)Follow37.6AccuracyFedMoon34.58435.36736.1536.933Oct 15, 2024Evaluation resultsMethodMethodLinksAccuracyCommunication Cost (GB)Computation Cost (TFLOPs)FedMoonC=5, Update Protocol=F...C=5, Update Protocol=FedPart, Backbone=ResNet-82024.1037.66.8816.1FedAvgC=2, Update Protocol=F...C=2, Update Protocol=FedPart, Backbone=ResNet-82024.1034.82.756.44FedProxC=2, Update Protocol=F...C=2, Update Protocol=FedPart, Backbone=ResNet-82024.1034.72.756.44