HQCLIP
Collection
7 items
•
Updated
Pretraining HQ-CLIP-B-16 on VLM-150M.
| Dataset | Performance |
|---|---|
| ImageNet 1k | 0.70556 |
| ImageNet V2 | 0.6308 |
| ImageNet-A | 0.391067 |
| ImageNet-O | 0.4295 |
| ImageNet-R | 0.801367 |
| ImageNet Sketch | 0.573189 |
| ObjectNet | 0.606439 |
| IN-shifts | 0.57206 |
| VTAB | 0.575571 |
| MSCOCO | 0.521573 |
| Flickr30k | 0.7786 |
| WinoGAViL | 0.528097 |
| Retrieval | 0.609423 |
| Avg. | 0.585715 |
- Downloads last month
- 93
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support