ViT
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Processing Inference | ViT (Vision Transformer) | Average Latency (ms)3.81 | 16 | |
| End-to-end inference tuning | ViT | Tuning Time (s)93.9 | 9 | |
| Image Classification | ViT-Base | Top-1 Accuracy31.68 | 3 | |
| Adversarial Attack | ViT Adversarially Trained | Attack Success Rate46.3 | 3 | |
| Vision | ViT-Base | Peak Performance Score2.05 | 1 |
Showing 5 of 5 rows