Image Classification on CIFAR-10 ρ=0.6 (test)
74.23AccuracyHARMONY
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HARMONYBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 74.23 | 68.3 | |
| SplitGPBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.22026.05 | 70.95 | 40.85 | |
| SplitGP + FTBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.2, Evaluation Protocol (Fine-tuning)=on-device fine-tuning2026.05 | 68.36 | 72.93 | |
| Multi-exit SFLBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 61.9 | 60.38 | |
| SplitGPBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=2, Personalization Intensity (λ)=0.92026.05 | 59.96 | 70.71 | |
| SplitFedBackbone=VGG-11, Number of Clients (N)=50, Data Partitioning (Shard)=22026.05 | 45.46 | 167.68 |