Image Classification on CIFAR-100 10 clients, heterog. (test)
46.6Test AccuracyIVON-ADMM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| IVON-ADMMBackbone=ResNet-20, Communication rounds=1002025.06 | 46.6 | 2.2 | |
| IVON-ADMM@mgBackbone=ResNet-20, Communication rounds=1002025.06 | 46.5 | 2.3 | |
| IVON-ADMM@mgBackbone=ResNet-20, Communication rounds=502025.06 | 46.2 | 2.1 | |
| IVON-ADMMBackbone=ResNet-20, Communication rounds=502025.06 | 46.2 | 2.1 | |
| FedAvg/FedProxBackbone=ResNet-20, Communication rounds=502025.06 | 40.4 | 2.6 | |
| FedLapBackbone=ResNet-20, Communication rounds=502025.06 | 40.4 | 3 | |
| FedLapBackbone=ResNet-20, Communication rounds=252025.06 | 40.1 | 2.9 | |
| IVON-ADMM@mgBackbone=ResNet-20, Communication rounds=252025.06 | 40 | 2.3 | |
| IVON-ADMMBackbone=ResNet-20, Communication rounds=252025.06 | 40 | 2.3 | |
| FedAvg/FedProxBackbone=ResNet-20, Communication rounds=1002025.06 | 39.8 | 3.4 | |
| FedLapBackbone=ResNet-20, Communication rounds=1002025.06 | 39.7 | 3.2 | |
| FedDynBackbone=ResNet-20, Communication rounds=1002025.06 | 39.6 | 3.2 | |
| FedDynBackbone=ResNet-20, Communication rounds=502025.06 | 39.2 | 3.2 | |
| FedDynBackbone=ResNet-20, Communication rounds=252025.06 | 38.4 | 3.5 | |
| FedAvg/FedProxBackbone=ResNet-20, Communication rounds=252025.06 | 37.9 | 2.4 |