Image Classification on CIFAR-10N (Accuracy and Training Time)
96.75AccuracySAM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SAMModel=ViT-B/16, Evaluation Protocol=Fine-tuning2025.10 | 96.75 | 656.7 | |
| fSGLDModel=ViT-B/16, Evaluation Protocol=Fine-tuning2025.10 | 96.45 | 345.8 | |
| ASAMModel=ViT-B/16, Evaluation Protocol=Fine-tuning2025.10 | 96.25 | 662.5 | |
| AdamWModel=ViT-B/16, Evaluation Protocol=Fine-tuning2025.10 | 95.57 | 344.5 | |
| SGDModel=ViT-B/16, Evaluation Protocol=Fine-tuning2025.10 | 94.64 | 343.2 | |
| ASAMModel=ResNet-34, Wall-clock time per iteration (s/epoch)=41.42025.10 | 91.73 | — | |
| SAMModel=ResNet-34, Wall-clock time per iteration (s/epoch)=41.32025.10 | 91.53 | — | |
| fSGLDModel=ResNet-34, Wall-clock time per iteration (s/epoch)=23.72025.10 | 91.37 | — | |
| ASAMModel=ResNet-50, Wall-clock time per iteration (s/epoch)=60.92025.10 | 91.25 | — | |
| SAMModel=ResNet-50, Wall-clock time per iteration (s/epoch)=60.72025.10 | 90.88 | — | |
| fSGLDModel=ResNet-50, Wall-clock time per iteration (s/epoch)=34.12025.10 | 90.86 | — | |
| SGDModel=ResNet-50, Wall-clock time per iteration (s/epoch)=31.92025.10 | 89.41 | — | |
| SGDModel=ResNet-34, Wall-clock time per iteration (s/epoch)=22.02025.10 | 89.31 | — | |
| AdamWModel=ResNet-50, Wall-clock time per iteration (s/epoch)=32.32025.10 | 89.26 | — | |
| AdamWModel=ResNet-34, Wall-clock time per iteration (s/epoch)=22.52025.10 | 89.25 | — |