mage
mage copied to clipboard
LTH14
A PyTorch implementation of MAGE: MAsked Generative Encoder to Unify Representation Learning and Image Synthesis
Hi！ When I read your source code, I found you set vocab_size = self.codebook_size + 1000 + 1 in token embbeding stage. Why not directly set vocal_size=self.codebook_size? What does the...
I want to mask part of the image and then rebuild it, not randomly.
Unexpected key(s) in state_dict: "head.weight", "head.bias", "fc_norm.weight", "fc_norm.bias". size mismatch for pos_embed: copying a param with shape torch.Size([1, 197, 1024]) from checkpoint, the shape in current model is torch.Size([1, 257,...
Dear author, Thanks for sharing the code. I am greatly interested in your work. I have a question for you and would like your reply. In the second stage, you...
Your work is fantastic and makes my eyes light up. I have done some experiments in MAE to improve the reconstruction of high-frequency features, but the results are mediocre. MAGE's...
Dear authors, I am super interested in your brilliant work! And I am very curious how image Inpainting has been achieved via such a framework. Different from previous MAE, MAGE...
Thank you very much for your sharing, I have questions. I'm trying to generate an image from a different version of the original input image. But the resulting image has...
Hi, thanks for sharing the Pytorch implementation! I am curious about how you select the stats for varied masking ratios. In the paper, you mentioned 'a truncated Gaussian distribution centered...
Hello, following your suggestion yesterday, I studied another one of your articles, MAGE. However, I'm curious about something that might not have been clearly explained. Figure 3 shows the unmasked...
