Skip to content
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

dataset attribute should not be set after SafeDataLoader is initialized #22

Open
JosephGatto opened this issue Nov 10, 2019 · 5 comments

Comments

@JosephGatto
Copy link

I have my own custom DataSet class. When I wrap it in nc.SafeDataset it runs with no error. When I then run loader = nc.SafeDataLoader(nc.SafeDataset(my_dataset)) it breaks and I get this error: dataset attribute should not be set after SafeDataLoader is initialized. Is this because I am not using a supported pytorch version?

@bhavika
Copy link

bhavika commented Nov 19, 2019

I'm seeing the same issue on PyTorch 1.3.0. Interested in knowing whether this is due to the version number or something else. I'm also using my own Dataset class.

@bhavika
Copy link

bhavika commented Nov 19, 2019

I'm seeing the same issue on PyTorch 1.3.0. Interested in knowing whether this is due to the version number or something else. I'm also using my own Dataset class.

Just adding here that I downgraded PyTorch to 1.2.0 and stopped seeing the error mentioned above.

@aRI0U
Copy link

aRI0U commented May 22, 2020

Commenting the line

self.dataset = _OriginalDataset(self.safe_dataset)

in nonechucks/dataloader.py seems to solve the issue (in PyTorch 1.5.0) in the basic example I tried on. However I assume that it breaks something else even if I don't know what

@chexenia
Copy link

any news on fixing it? I still have the same problem with pytorch 1.5.1

@ms-lilibeth
Copy link

I experience the same problem. I seems that @msamogh abandoned this project development

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants