Image Classification on CIFAR-100 (Directional Accuracy Metrics)
66.84Accuracy (Dir-0.6)FedACT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FedACTBackbone=ResNet-18, Communication Rounds=3002026.07 | 66.84 | 66.62 | 64.42 | |
| FedAdamWBackbone=ResNet-18, Communication Rounds=3002026.07 | 65.43 | 65.86 | 63.24 | |
| SCAFFOLDBackbone=ResNet-18, Communication Rounds=3002026.07 | 64.96 | 64.58 | 61.43 | |
| FedAvgBackbone=ResNet-18, Communication Rounds=3002026.07 | 64.93 | 64.5 | 60.88 | |
| FedProxBackbone=ResNet-18, Communication Rounds=3002026.07 | 64.54 | 64.28 | 60.33 | |
| FedLADABackbone=ResNet-18, Communication Rounds=3002026.07 | 64.3 | 63.95 | 61.31 | |
| LocalAdamWBackbone=ResNet-18, Communication Rounds=3002026.07 | 64.23 | 63.75 | 59.78 | |
| LocalAdamBackbone=ResNet-18, Communication Rounds=3002026.07 | 63.95 | 63.7 | 59.69 | |
| FedAdamBackbone=ResNet-18, Communication Rounds=3002026.07 | 62.55 | 62.21 | 57.82 |