Shujun WANG
Shujun WANG
ok. Thanks a lot.
So for all testing data, they have the same threshold right? Is it a constant that you need to find when training the model?
The deepall results are inheriated from paper: Domain Generalization viaModel-Agnostic Learning of Semantic Features. We remove the specific components to generate deep all + triplet, deep all + jigsaw results....
The results reported in the paper are the average of the first three runs. Before release this code, we also have reproduced the method with this code which results are...
Is this adversarial network implementation available yet? Thanks!
I still cannot understand how to fix this issue?? Could you tell me the details steps?? Thanks a lot.
Here for the code also! Great paper!
May I know how to specify the version of PyG? According to the official install document from the PyG website, it is `pip install pyg-lib torch-scatter torch-sparse -f https://data.pyg.org/whl/torch-1.13.0+cu116.html `...
....me too