Decentralized Distributed Learning on Nonconvex Functions
1Communication CostPD-SGD
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PD-SGDCompressed Communication=false, Adaptive Gradient=false2026.04 | 1 | 1 | — | — | |
| LSGTCompressed Communication=false, Adaptive Gradient=false2026.04 | 1 | 1 | — | — | |
| DFedAvgMCompressed Communication=true, Adaptive Gradient=false2026.04 | 1 | — | — | — | |
| SQuARM-SGDCompressed Communication=true, Adaptive Gradient=false2026.04 | 1 | 1 | — | — | |
| LM-DFLCompressed Communication=true, Adaptive Gradient=false2026.04 | 1 | 1 | — | — | |
| LoDAdaCCompressed Communication=true, Adaptive Gradient=true2026.04 | 1 | 1 | — | — |