Continual Learning on CIFAR100 Coarse
92.64Avg Per-Task AccuracyLITELORA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LITELORATask Order=Order 22026.06 | 92.64 | 5.14 | 8 | |
| SD-LoRATask Order=Order 22026.06 | 92.54 | 4.93 | 10 | |
| LITELORATask Order=Order 12026.06 | 91.81 | 4.66 | 5 | |
| SD-LoRATask Order=Order 12026.06 | 91.71 | 5.92 | 10 | |
| LITELORATask Order=Order 32026.06 | 91.41 | 6.09 | 7 | |
| SD-LoRATask Order=Order 32026.06 | 91.27 | 5.96 | 10 | |
| Model ZooReplay=true, Single Epoch=false2021.06 | 84.27 | — | — | |
| Multi-Head (multi-task)2021.06 | 83.19 | — | — | |
| IsolatedReplay=false, Single Epoch=false2021.06 | 80.72 | — | — | |
| RMNReplay=false, Single Epoch=false2021.06 | 80.01 | — | — | |
| Model Zoo-small (10% replay)Replay=true, Single Epoch=false2021.06 | 77.18 | — | — | |
| Model Zoo-smallReplay=true, Single Epoch=false2021.06 | 73.72 | — | — | |
| Isolated-smallReplay=false, Single Epoch=false2021.06 | 69.07 | — | — | |
| APDReplay=false, Single Epoch=false2021.06 | 56.81 | — | — | |
| FedWeITReplay=false, Single Epoch=false2021.06 | 55.16 | — | — |