Federated Learning Classification on Tiny-ImageNet non-iid Dirichlet 0.6 (test)
0.3607AccuracyFedDC+MAN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FedDC+MANRegularizer=MAN2025.12 | 0.3607 | 66 | |
| FedSpeed+MANRegularizer=MAN2025.12 | 0.3549 | 78 | |
| FedSAM+MANRegularizer=MAN2025.12 | 0.326 | 87 | |
| FedDyn+MANRegularizer=MAN2025.12 | 0.3244 | 110 | |
| FedASAM+MANRegularizer=MAN2025.12 | 0.3241 | 136 | |
| FedDCRegularizer=None2025.12 | 0.3142 | 193 | |
| FedSpeedRegularizer=None2025.12 | 0.311 | 228 | |
| FedDynRegularizer=None2025.12 | 0.2899 | 308 | |
| FedAvg+MANRegularizer=MAN2025.12 | 0.289 | 194 | |
| FedASAMRegularizer=None2025.12 | 0.2705 | 500 | |
| FedSAMRegularizer=None2025.12 | 0.2604 | 500 | |
| FedAvgRegularizer=None2025.12 | 0.2441 | 500 |