nSircombe

Fantastic! Thanks @igorsafo.

Hi @vpirogov, I can take a look, but I've not looked at Conda packaging for oneDNN - it doesn't look like the failure should be specific to that though? We...

Hi @vpirogov, Sorry for the slow response, this slipped off my to-do list. I'm able to reproduce the failure with `DNNL_CPU_RUNTIME=TBB` builds, an a number of other segfaults including `test_eltwise`...

Hi @ZJUFangzh I can cover what happens for the Compute Library kernels. _Note: for more information on how Compute Library is integrated into oneDNN see this RFC: https://github.com/oneapi-src/oneDNN/pull/795/files#diff-c03394b05656627a041b9bb8ef2e6cd10952058216bb9f0f96bfdf53a5a95b86 and subsequent...

Hi @ZJUFangzh, > The second problem is, when we run Resnet on AARCH64, the memory continues to grow, and when I set OMP_NUM_THREAD=1, the memory remains stable. ...that's interesting. Would...

Hi @igorsafo, Thanks for the heads-up, we'll take a look. What is the timeframe for the oneDNN 3.0 release?

Hi @igorsafo, It looks like supporting these changes with ACL would require changes to the ACL API, so as it stands this would introduce problems for the current quantised GEMM...

Hi @malfet, That's interesting, I have not hit that issue myself - I'd be very interested to know how you get on. If we can reproduce then I'm happy to...

HI @malfet, Your 'smoketest' works for me on my OpenMP-ed from-src build. I notice that the `torch.__config__` on the released whl shows `USE_EIGEN_FOR_BLAS=ON` - this is not the case for...

Here's a reproducer for the "Detect OpenMP Loop and this application may hang." warning. ``` import torch import urllib import torchvision.models as models from PIL import Image from torchvision import...