Efficient Vits On Imagenet 1K With Deit S
Metriken
GFLOPs
Top 1 Accuracy
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
Vergleichstabelle
| Modellname | GFLOPs | Top 1 Accuracy |
|---|---|---|
| scalable-visual-transformers-with | 2.7 | 78.3 |
| dynamicvit-efficient-vision-transformers-with | 2.9 | 79.3 |
| learned-thresholds-token-merging-and-pruning | 2.3 | 78.6 |
| not-all-patches-are-what-you-need-expediting | 3.5 | 79.8 |
| beyond-attentive-tokens-incorporating-token | 3.0 | 79.6 |
| multi-criteria-token-fusion-with-one-step | 2.4 | 79.9 |
| beyond-attentive-tokens-incorporating-token | 2.6 | 79.3 |
| ia-red-2-interpretability-aware-redundancy | 3.2 | 79.1 |
| adaptive-sparse-vit-towards-learnable | 2.3 | 78.7 |
| chasing-sparsity-in-vision-transformers-an | 3.2 | 79.2 |
| beyond-attentive-tokens-incorporating-token | 1.6 | 76.4 |
| patch-slimming-for-efficient-vision | 2.4 | 79.5 |
| learned-thresholds-token-merging-and-pruning | 3.8 | 79.8 |
| pruning-self-attentions-into-convolutional | 3.3 | 78.3 |
| diffrate-differentiable-compression-rate-for | 2.9 | 79.8 |
| adaptive-sparse-vit-towards-learnable | 3.0 | 79.6 |
| training-data-efficient-image-transformers | 4.6 | 79.8 |
| not-all-patches-are-what-you-need-expediting | 4.0 | 79.8 |
| not-all-patches-are-what-you-need-expediting | 2.6 | 78.9 |
| ppt-token-pruning-and-pooling-for-efficient | 2.9 | 79.8 |
| multi-criteria-token-fusion-with-one-step | 2.2 | 79.5 |
| token-merging-your-vit-but-faster | 3.4 | 79.7 |
| dynamicvit-efficient-vision-transformers-with | 4.0 | 79.8 |
| beyond-attentive-tokens-incorporating-token | 2.3 | 79.0 |
| adavit-adaptive-tokens-for-efficient-vision | 3.6 | 78.6 |
| not-all-patches-are-what-you-need-expediting | 2.3 | 78.5 |
| beyond-attentive-tokens-incorporating-token | 1.8 | 77.8 |
| spvit-enabling-faster-vision-transformers-via | 2.6 | 79.3 |
| patch-slimming-for-efficient-vision | 2.6 | 79.4 |
| not-all-patches-are-what-you-need-expediting | 3.0 | 79.5 |
| learned-thresholds-token-merging-and-pruning | 3.0 | 79.6 |
| ats-adaptive-token-sampling-for-efficient | 2.9 | 79.7 |
| evo-vit-slow-fast-token-evolution-for-dynamic | 3.0 | 79.4 |
| joint-token-pruning-and-squeezing-towards | 3.0 | 79.7 |
| joint-token-pruning-and-squeezing-towards | 3.0 | 80.1 |
| token-merging-your-vit-but-faster | 2.7 | 79.4 |
| token-merging-your-vit-but-faster | 2.3 | 79.1 |
| dynamicvit-efficient-vision-transformers-with | 3.4 | 79.8 |
| beyond-attentive-tokens-incorporating-token | 2.0 | 78.6 |
| multi-criteria-token-fusion-with-one-step | 2.6 | 80.1 |
| spvit-enabling-faster-vision-transformers-via | 3.9 | 79.8 |