| Benchmark | Model | Rank | Results |
|---|---|---|---|
| image-classification-on-imagenet | FD (CLIP ViT-L-336) | #25 | Top 1 Accuracy: 89.0%Number of params: 307M |
| instance-segmentation-on-coco | FD-SwinV2-G | #4 | mask AP: 55.4 |
| object-detection-on-coco | FD-SwinV2-G | #10 | box mAP: 64.2 |
| semantic-segmentation-on-ade20k | FD-SwinV2-G | #9 | Validation mIoU: 61.4Params (M): 3000 |
| semantic-segmentation-on-ade20k-val | FD-SwinV2-G | #4 | mIoU: 61.4 |