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

Annoying warning in Pytorch 1.12. #45

Open
pglez82 opened this issue Jun 14, 2024 · 1 comment
Open

Annoying warning in Pytorch 1.12. #45

pglez82 opened this issue Jun 14, 2024 · 1 comment

Comments

@pglez82
Copy link

pglez82 commented Jun 14, 2024

Hi. I am using your library using PyTorch 1.10. It works great! One of my colleagues updated to PyTorch 1.12, and the library still works but, it fills the console with logs that are related with geotorch.

Variable.execution_engine.run_backward( # Calls into the C++ engine to run the backward pass /pathtomyenvironment/python3.9/site-packages/geotorch/so.py:83: UserWarning: An output with one or more elements was resized since it had shape [114, 17, 17], which does not match the required output shape [1, 114, 17, 17]. This behaviour is deprecated, and in a future PyTorch release outputs will not be resized unless they have zero elements. You can explicitly reuse an out tensor t by resizing it, inplace, to zero elements with t.resize(0). (Triggered internally at /opt/conda/conda-bld/pytorch_1659484775609/work/aten/src/ATen/native/Resize.cpp:24.)

I do not know if it is my fault, with this extra tensor dimension that it is complaining about, but the reality is that everything works and with PyTorch 1.10 no warning at all is shown.

@lezcano
Copy link
Owner

lezcano commented Jun 14, 2024

Hi Pablo!
I'm afraid I'm going to need a small reproducer where you hit this issue. For debugging, consider using https://pytorch.org/docs/stable/autograd.html#torch.autograd.detect_anomaly to debug this sort of issues.

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

2 participants