Zhenlin Xu

I also met this issue. It is weird that the issue arises in some function but not others that all calls the same visual encoder `self.ln_vision(self.visual_encoder(image))`. Does anyone know why?

@Hongben-fba Thanks for your interests. The current released version only has the pure segmentation model training. The registration and segmentation+registration code is not released yet.

Hi Haohan, thanks for sharing the lists and the source repo. You mentioned in the paper that JigenDG used a different split. What you uploaded are not what JigenDG used,...

Hi Haohan, I meet some issues when trying to replicate the results on PACS. It would be great if you could share your ideas on these questions. 1. Could you...

> Hi Zhenlin, sure, but it has been quite a while, I'm not sure I remember everything but I will give you as much information as possible. First of all,...

> Yes, we used the weights in the link. > I believe only this AlexNet architecture uses conv5. The ResNet in other experiments uses conv2. Thank you Haohan. Could you...

Thanks! Currently, I have tried feature_source = {conv1, conv2, conv3, conv5} X lambda={0.01, 0.001} but still cannot get performance close to what were reported in the paper. So far the...

> At line https://github.com/HaohanWang/PAR_experiments/blob/master/PACS/alexNet.py#L322, > it looks like I used 1e-5 for art and sketch as the test domain, but something else for the other settings. I checked the git...

> Yes, that's the most I can recall. My experiments on 1080Ti. The CIFAR experiments were run and documented by my collaborator though. Sure. I am only testing on PACS...

> Hi @wildphoton , we are working on the local support of this feature (insufficient bandwidth recently lol). But if you want, you could refer to [here](https://github.com/Psycoy/MixEval/blob/main/docs/evaluate_instructions.md#use-other-apis-for-model-parser) to use the...