SPViT: Enabling Faster Vision Transformers via Soft Token Pruning

Open paper
Benchmark Model Rank Results
efficient-vits-on-imagenet-1k-with-deit-sSPViT (3.9G)#7Top 1 Accuracy: 79.8GFLOPs: 3.9
efficient-vits-on-imagenet-1k-with-deit-sSPViT (2.6G)#24Top 1 Accuracy: 79.3GFLOPs: 2.6
efficient-vits-on-imagenet-1k-with-deit-tSPViT (1.0G)#7Top 1 Accuracy: 72.2GFLOPs: 1.0
efficient-vits-on-imagenet-1k-with-deit-tSPViT (0.9G)#8Top 1 Accuracy: 72.1GFLOPs: 0.9
efficient-vits-on-imagenet-1k-with-lv-vit-sSPViT#6Top 1 Accuracy: 83.1GFLOPs: 4.3

Rank counts only results with a code link.