johnnv commited on
Commit
0905914
1 Parent(s): 3ebdd26

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -12,8 +12,8 @@ vit_model = VisionTransformer.from_config('vit_s/32', pretrained=True)
12
  ...
13
  ```
14
 
15
- Original weights from https://github.com/google-research/vision_transformer: This weight is based on the
16
- [Original ViT_S/32 pretrained on imagenet21k](https://storage.googleapis.com/vit_models/augreg/S_32-i21k-300ep-lr_0.001-aug_none-wd_0.1-do_0.0-sd_0.0.npz)
17
 
18
  Weights converted to PyTorch for Kornia ViT implementation (by [@gau-nernst](https://github.com/gau-nernst) in [kornia/kornia#2786](https://github.com/kornia/kornia/pull/2786#discussion_r1482339811))
19
  <details>
 
12
  ...
13
  ```
14
 
15
+ Original weights from [AugReg](https://arxiv.org/abs/2106.10270) as recommended by [google research vision transformer repo](https://github.com/google-research/vision_transformer): This weight is based on the
16
+ [AugReg ViT_S/32 pretrained on imagenet21k](https://storage.googleapis.com/vit_models/augreg/S_32-i21k-300ep-lr_0.001-aug_none-wd_0.1-do_0.0-sd_0.0.npz)
17
 
18
  Weights converted to PyTorch for Kornia ViT implementation (by [@gau-nernst](https://github.com/gau-nernst) in [kornia/kornia#2786](https://github.com/kornia/kornia/pull/2786#discussion_r1482339811))
19
  <details>