Hugging Face's logo

laion
/
CLIP-convnext_large_d_320.laion2B-s29B-b131K-ft

Zero-Shot Image Classification
OpenCLIP
TensorBoard
clip
Model card Files Metrics Community
rwightman HF staff commited on
Commit
dc8fa1e
·
1 Parent(s): e07fdf6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -35,8 +35,8 @@ At 320x320, the ConvNext-Large-D is significantly more efficient than the L/14 m
35
  | Model | Dataset | Resolution | AugReg | Top-1 ImageNet Zero-Shot (%) |
36
  | ----- | ------- | ---------- | ------------ | --------- |
37
  | [convnext_large_d.laion2b_s26b_b102k-augreg](https://huggingface.co/laion/CLIP-convnext_large_d.laion2B-s26B-b102K-augreg) | LAION-2B | 256x256 | RRC (0.33, 1.0), RE (0.35), SD (0.1), D(0.1) | 75.9 |
38
- | [convnext_large_d_320.laion2b_s29b_b131k-augreg-ft](https://huggingface.co/laion/CLIP-convnext_large_d_320.laion2B-s29B-b131K-augreg-ft) | LAION-2B | 320x320 | RRC (0.5, 1.0), RE (0.4), SD (0.1), D(0.0) | 76.6 |
39
- | [convnext_large_d_320.laion2b_s29b_b131k-augreg-ft-soup](https://huggingface.co/laion/CLIP-convnext_large_d_320.laion2B-s29B-b131K-augreg-ft-soup) | LAION-2B | 320x320 | RRC (0.5, 1.0), RE (0.4), SD (0.1), D(0.0) | 76.9 |
40
 
41
  RRC = Random Resize Crop (crop pcts), RE = Random Erasing (prob), SD = Stochastic Depth (prob) -- image tower only, D = Dropout (prob) -- image tower head only
42
 
 
35
  | Model | Dataset | Resolution | AugReg | Top-1 ImageNet Zero-Shot (%) |
36
  | ----- | ------- | ---------- | ------------ | --------- |
37
  | [convnext_large_d.laion2b_s26b_b102k-augreg](https://huggingface.co/laion/CLIP-convnext_large_d.laion2B-s26B-b102K-augreg) | LAION-2B | 256x256 | RRC (0.33, 1.0), RE (0.35), SD (0.1), D(0.1) | 75.9 |
38
+ | [convnext_large_d_320.laion2b_s29b_b131k-ft](https://huggingface.co/laion/CLIP-convnext_large_d_320.laion2B-s29B-b131K-ft) | LAION-2B | 320x320 | RRC (0.5, 1.0), RE (0.4), SD (0.1), D(0.0) | 76.6 |
39
+ | [convnext_large_d_320.laion2b_s29b_b131k-ft-soup](https://huggingface.co/laion/CLIP-convnext_large_d_320.laion2B-s29B-b131K-ft-soup) | LAION-2B | 320x320 | RRC (0.5, 1.0), RE (0.4), SD (0.1), D(0.0) | 76.9 |
40
 
41
  RRC = Random Resize Crop (crop pcts), RE = Random Erasing (prob), SD = Stochastic Depth (prob) -- image tower only, D = Dropout (prob) -- image tower head only
42