The code for contrastive leanring is included in the current codebase or not?

Open gaopengpjlab opened this issue 3 years ago • 4 comments

As the title says, I can not find the contrastive learning loss in your codebase.

Feb 14 '23 05:02 gaopengpjlab

I just updated the README to include pre-trained MAGE-C checkpoints converted from JAX. We are still wrapping up the PyTorch code for the contrastive loss. We will endeavor to release that part soon.

Feb 14 '23 06:02 LTH14

Can you provide the performance comparison between models trained by MIM and MIM+CL?

Feb 14 '23 06:02 gaopengpjlab

Please refer to this paper for that comparison. We also have the comparison between MAGE and MAGE-C in Table 2 and Table 13 in our paper.

Feb 14 '23 06:02 LTH14

@LTH14 Is MAGE-C ready?

Jul 10 '24 04:07 densechen