904763189cy

https://www.cs.cmu.edu/~clean-fid/stats/ffhq_clean_train_1024.npz 这个网页是404 在https://www.cs.cmu.edu/~clean-fid/stats没有ffhq_clean_train_1024.npz，但是有ffhq_clean_trainval_1024.npz，请问一下，代码本来是想下载ffhq_clean_trainval_1024.npz吗？

No arcface_resnet18.pth found in arcface-pytorch

2

Could you provide a download link for the pretrained model？ Thanks

delta = torch.zeros(self.lq.shape[0], self.lq.shape[1], int(self.lq.shape[2]//self.opt['noise_scale']), int(self.lq.shape[3]//self.opt['noise_scale'])).to(self.device) Why use ‘//self.opt[‘noise_scale’]’?

After running this code：output, attributes = restoration_inference(model, os.path.join(args.file_path), return_attributes=True) output and attributes are the same. Why return the same two data? Code running result： >>>>>>>>>>>>>>attributes>>>>>>>>>>>>>>>>>> tensor([[0.9644, 0.3345, 0.9995, 0.0357, 0.2766,...