Image Classification on ImageNet (val) (Accuracy and Norm. AxM Power)
81.83AccuracyAccMul
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AccMulModel=Swin-S, Precision=8-bit2026.05 | 81.83 | — | |
| TRAMModel=Swin-S, λ=12026.05 | 79.54 | 82.53 | |
| AccMulModel=DeiT-S, Precision=8-bit2026.05 | 79.34 | — | |
| Evo_1KV9Model=Swin-S2026.05 | 79.25 | 93.17 | |
| TRAMModel=Swin-S, λ=1002026.05 | 79.15 | 66.08 | |
| TRAMModel=DeiT-S, λ=12026.05 | 76.78 | 82.53 | |
| Evo_1L2HModel=Swin-S2026.05 | 76.64 | 73.04 | |
| TRAMModel=DeiT-S, λ=1002026.05 | 76.31 | 66.08 | |
| Evo_1KV9Model=DeiT-S2026.05 | 70.16 | 93.17 | |
| Evo_1L2HModel=DeiT-S2026.05 | 67.01 | 73.04 |