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

add patch to fix flaky test optim test in PyTorch 1.12.1 w/ foss/2022a + CUDA 11.7.0 #17726

Merged

Conversation

VRehnberg
Copy link
Contributor

(created using eb --new-pr)

@Micket Micket added the bug fix label Apr 14, 2023
@Micket Micket added this to the next release (4.7.2) milestone Apr 14, 2023
@Micket
Copy link
Contributor

Micket commented Apr 14, 2023

Test report by @Micket
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
alvis3-19 - Linux Rocky Linux 8.6, x86_64, Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz, 4 x NVIDIA NVIDIA A100-SXM4-40GB, 520.61.05, Python 3.6.8
See https://gist.github.com/Micket/ddb4cb16374bc7b6bd84b3f946627a8e for a full test report.

@boegel
Copy link
Member

boegel commented Apr 15, 2023

@VRehnberg Should we also add this patch to the other easyconfigs for PyTorch 1.12.1?

PyTorch-1.12.1-foss-2021a-CUDA-11.3.1.eb
PyTorch-1.12.1-foss-2021a.eb
PyTorch-1.12.1-foss-2021b-CUDA-11.5.2.eb
PyTorch-1.12.1-foss-2021b.eb
PyTorch-1.12.1-foss-2022a-CUDA-11.7.0.eb
PyTorch-1.12.1-foss-2022a.eb

@Micket
Copy link
Contributor

Micket commented Apr 15, 2023

I see no reason not to. Lets merge this first, it takes a million billion years for the test reports as usual

@branfosj
Copy link
Member

I see no reason not to. Lets merge this first, it takes a million billion years for the test reports as usual

I agree. This is changing one test, so I am happy with the one successful test report. I'll follow up with PRs for the other easyconfigs.

@branfosj
Copy link
Member

Going in, thanks @VRehnberg!

@boegel
Copy link
Member

boegel commented Apr 15, 2023

I see no reason not to. Lets merge this first, it takes a million billion years for the test reports as usual

I agree that doing that in a single PR would have been bloody painful w.r.t. testing, that's not was I was suggesting.

Doing it in separate PRs like @branfosj is the other extreme, but it doesn't hurt either. Good for PR stats, that's for sure. ;P

@boegel boegel changed the title PyTorch-1.12.1-foss-2022a-CUDA-11.7.0 fix flaky test optim test add patch to fix flaky test optim test in PyTorch 1.12.1 w/ foss/2022a + CUDA 11.7.0 Apr 15, 2023
@branfosj
Copy link
Member

See #17737 - this change looks to be causing issues with other tests in test optim.

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

Successfully merging this pull request may close these issues.

None yet

4 participants