-
Notifications
You must be signed in to change notification settings - Fork 14
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cannot be reproduced #8
Comments
You can check the inputs to the model. I'll update a few times later on how to use this training code. |
Is class.json automatically generated in preprocessing? I didn't see such a file after running the code for preprocessing. And I got FileNotFound error when running train_video.py without the file. |
I didn't see this file either |
see closed issue#6 |
Same problem here. The training behavior seems not normal. The training loss is low (ranges from -10000 to -2000 after a few steps). Any comments? |
@dengxl0520 @Zong-Liang |
The issue labeled "issue 6" has been closed. Could you please inform me how to proceed with this? |
it's from the SAMUS repo,you can refer to it,but the class.json is not the key problem here, i've abandoned reproduction(My abilities are not enough),good luck to U. |
No,i've abandoned,wasted too much time here~ |
Hello, have you resolved the issue with the |
Thank you very much for your answer
Thank you very much for your answer |
我已经解决该问题了,猜测原因是作者未给出专门用于处理camus数据集的代码部分,因此导致camus的mask张量中出现255,正确的mask张量中的元素应该是非0即1的,因此需要在data_us文件中的第501行加入代码:mask[mask == 255] = 1。 |
Dear author, thanks for your great work. I encountered the following issues when trying to reproduce the paper:
However, I am unable to train normally:
I'm not sure why this is happening. Could you please help me? I'd greatly appreciate your help. Thank you so much.
The text was updated successfully, but these errors were encountered: