-
Notifications
You must be signed in to change notification settings - Fork 24.1k
Floating Point Exception on PyTorch nightlies #17029
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
Comments
This is fixed in the latest mkl-dnn https://github.com/intel/mkl-dnn/blob/b88908909240bdab10f1cb7c256caa264fc15067/src/cpu/xbyak/xbyak_util.h#L216, but not in the mkl-dnn hash pytorch is using https://github.com/intel/mkl-dnn/blob/830a10059a018cd2634d94195140cf2d8790a75a/src/cpu/xbyak/xbyak_util.h#L186. |
#16183 is related |
@mingfeima can you help get this patch into a ideep/mkl-dnn release that pytorch uses. For v1.0.1, I manually patched the source before building binaries. |
@soumith will do. |
Thank you. |
🐛 Bug
We are seeing a floating point exception in the nightly containers but not in v1.0.1 container. I believe there is a mkl-dnn patch that needs to be on master?
@soumith
The text was updated successfully, but these errors were encountered: