Federated Learning Classification on Tiny-ImageNet non-iid Dirichlet 0.3 (test)
35.7AccuracyFedDC+MAN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FedDC+MANRegularizer=MAN2025.12 | 35.7 | 73 | |
| FedSpeed+MANRegularizer=MAN2025.12 | 34.32 | 108 | |
| FedASAM+MANRegularizer=MAN2025.12 | 32.5 | 140 | |
| FedSAM+MANRegularizer=MAN2025.12 | 32.16 | 104 | |
| FedDyn+MANRegularizer=MAN2025.12 | 32 | 132 | |
| FedDCRegularizer=None2025.12 | 31.44 | 170 | |
| FedSpeedRegularizer=None2025.12 | 31.12 | 211 | |
| FedDynRegularizer=None2025.12 | 29.23 | 295 | |
| FedAvg+MANRegularizer=MAN2025.12 | 28.09 | 437 | |
| FedASAMRegularizer=None2025.12 | 27.5 | 500 | |
| FedSAMRegularizer=None2025.12 | 26.23 | 500 | |
| FedAvgRegularizer=None2025.12 | 25.35 | 500 |