can not reproduce the results

[kristine-li]

Hi, Thanks for your great work. I enjoyed reading your paper! And really thanks for sharing your code and model checkpoint.
When running main_stablenet.py, I experience no reproducibility of the results。
Example：

Dataset：PACS（unbalanced(5:1:1) | PACS | C,S,P | art_painting | ResNet18） ，and we also use the dataset split files you provided。
The best results which we reproduce：

Do you have a clue why that is the case？
Thanks in advance!

[xxgege]

Hi,
Thank you for your interest in our work! As mentioned in readme, many hyper-parameters can affect the performance (such as lrbl, epochb, lambdap), which can vary among different tasks/environments/software/hardware/random seeds, and thus carefully tunning is required. You may need to tune them according to your environment and software.
We are sorry for this problem and trying to address it in future work.

[xioahua]

Hi, Thanks for your great work. I enjoyed reading your paper! And really thanks for sharing your code and model checkpoint. When running main_stablenet.py, I experience no reproducibility of the results。 Example：

Dataset：PACS（unbalanced(5:1:1) | PACS | C,S,P | art_painting | ResNet18） ，and we also use the dataset split files you provided。 The best results which we reproduce：

Do you have a clue why that is the case？ Thanks in advance!

Hi, could you provide the same dataset split files for me? thank you.

[1424163067]

Hi, Thanks for your great work. I enjoyed reading your paper! And really thanks for sharing your code and model checkpoint.
When running main_stablenet.py, some questions have arisen that I would like to ask you.
Example:

I have created a directory and put the dataset under the directory as required by the error, but still report this error.


Do you have a clue why that is the case？
Thanks in advance!

[allenwang2333]

May be try change default path in ops/config.py of data directory