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.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

RuntimeError caused by batchnorm #86790

Closed
CaoE opened this issue Oct 12, 2022 · 4 comments
Closed

RuntimeError caused by batchnorm #86790

CaoE opened this issue Oct 12, 2022 · 4 comments
Labels
module: norms and normalization triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module

Comments

@CaoE
Copy link
Collaborator

CaoE commented Oct 12, 2022

馃悰 Describe the bug

The model TTS will crash due to the issue: when input of BN is not contiguous and the data type of input is different with that of parameters, BN will raise error RuntimeError: !needs_dynamic_casting<func_t>::check(iter) INTERNAL ASSERT FAILED at "xxx/pytorch/aten/src/ATen/native/cpu/Loops.h":311, please report a bug to PyTorch.

Versions

PyTorch master and PyTorch 1.13
python 3.8.5
@CaoE CaoE added the module: crash Problem manifests as a hard crash, as opposed to a RuntimeError label Oct 12, 2022
@CaoE CaoE mentioned this issue Oct 12, 2022
Closed
@kobistar
Copy link

@CaoE can I claim this issue?

@CaoE
Copy link
Collaborator Author

CaoE commented Oct 12, 2022

Thank you @kobistar and there is a PR #84410 to fix the issue.

@samdow samdow added triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module module: norms and normalization labels Oct 12, 2022
@malfet malfet removed the module: crash Problem manifests as a hard crash, as opposed to a RuntimeError label Oct 12, 2022
@malfet
Copy link
Contributor

malfet commented Oct 12, 2022

Removed module: crash label as this is a structured exception.
Even description of module: crash says that it should not be used for RuntimeErrors:

Problem manifests as a hard crash, as opposed to a RuntimeError

@malfet malfet linked a pull request Oct 13, 2022 that will close this issue
Make the data types of output and input consistenst for batchnorm #84410
Closed
@malfet
Copy link
Contributor

malfet commented Oct 13, 2022

#84410 should have been marked as closing the issue

@malfet malfet closed this as completed Oct 13, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
module: norms and normalization triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Make the data types of output and input consistenst for batchnorm CaoE/pytorch
4 participants
@malfet @samdow @CaoE @kobistar