askerlee

Thanks for providing ethics-wise accuracy. Seems the embeddings are particularly bad at differentiating asian faces.

EfficientNet-B4

可以参考这个report：https://arxiv.org/abs/2202.08994

The supported backbones are: CONFIGS = { 'ViT-B_16': configs.get_b16_config(), 'ViT-B_32': configs.get_b32_config(), 'ViT-L_16': configs.get_l16_config(), 'ViT-L_32': configs.get_l32_config(), 'ViT-H_14': configs.get_h14_config(), 'R50-ViT-B_16': configs.get_r50_b16_config(), 'R50-ViT-L_16': configs.get_r50_l16_config(), 'R50-ViT-L_32': configs.get_r50_l32_config(), 'testing': configs.get_testing(), } (This dict is in...

I've encountered the same problem. Would you like to share how you work around it? Thanks.

@x-CK-x I see. I did find a workaround later (with the help of O1, of course): ``` css = ''' .custom-gallery { height: 800px !important; width: 100%; margin: 10px auto;...

I just did git clone and it's good. Maybe you can check your internet connections and see if it's stable? rafter@ass ~> git clone https://github.com/askerlee/segtran/ Cloning into 'segtran'... remote: Enumerating...