Image Classification on CIFAR100-PAM
55.44AccuracyFedSAM
Evaluation results
| Method | Links | |
|---|---|---|
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Stochastic Weight Averaging=true2022.03 | 55.44 | |
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Stochastic Weight Averaging=true2022.03 | 54.67 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Stochastic Weight Averaging=true2022.03 | 54.29 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Stochastic Weight Averaging=true2022.03 | 54.25 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 53.5 | |
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 53.41 | |
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 53.1 | |
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 52.96 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Stochastic Weight Averaging=true2022.03 | 52.7 | |
| FedASAMBackbone=ResNet-18, Augmentation=Mixup, Local Epochs (E)=2, Number of clients=10, Stochastic Weight Averaging=true2022.03 | 52 | |
| FedSAMBackbone=ResNet-18, Augmentation=Cutout, Local Epochs (E)=1, Number of clients=20, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 51.92 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 51.1 | |
| FedSAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 50.71 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Stochastic Weight Averaging=true2022.03 | 50.28 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 49.31 | |
| FedASAMBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 47.21 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 47.03 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=20, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 46.6 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 46.51 | |
| FedAvgBackbone=ResNet-18, Augmentation=None, Local Epochs (E)=1, Number of clients=10, Training Rounds=10k, Stochastic Weight Averaging=false2022.03 | 45.83 | |
| FedAvgBackbone=ResNet-18, Augmentation=Cutout, Local Epochs (E)=1, Number of clients=20, Training Rounds=5k, Stochastic Weight Averaging=false2022.03 | 43.47 |