Image Classification on CIFAR-10 (test) (Accuracy, Sparsity, and Drop)
54.08SparsityMix-and-Match
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Mix-and-MatchBackbone=LeViT-384, Target Sparsity=~54%2026.03 | 54.08 | 97.54 | 0.21 | |
| MAGBackbone=LeViT-384, Target Sparsity=~54%2026.03 | 54 | 96.5 | 1.25 | |
| SNIPBackbone=LeViT-384, Target Sparsity=~54%2026.03 | 54 | 96.95 | 0.8 | |
| GraSPBackbone=LeViT-384, Target Sparsity=~54%2026.03 | 54 | 97.16 | 0.59 | |
| LPViTBackbone=LeViT-384, Target Sparsity=~54%2026.03 | 53.91 | 97.7 | 0.05 | |
| GETABackbone=LeViT-384, Target Sparsity=~54%2026.03 | 53.91 | 97.17 | 0.58 | |
| Mix-and-MatchBackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 44.78 | 87.64 | 1.5 | |
| MAGBackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 44 | 86.35 | 2.79 | |
| SNIPBackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 44 | 85.85 | 3.29 | |
| GraSPBackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 44 | 85.48 | 3.66 | |
| LPViTBackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 43.88 | 86.7 | 2.44 | |
| GETABackbone=Swin-Tiny, Target Sparsity=~45%2026.03 | 43.88 | 84.23 | 4.91 |