{devel}[foss/2021a] PyTorch v1.10.0 w/ Python 3.9.5 (CPU-only)

[boegel]

(created using eb --new-pr)

[boegel]

@boegelbot please test @ generoso
CORE_CNT=16

[boegelbot]

@boegel: Request for testing this PR well received on login1

PR test command 'EB_PR=14460 EB_ARGS= /opt/software/slurm/bin/sbatch --job-name test_PR_14460 --ntasks="16" ~/boegelbot/eb_from_pr_upload_generoso.sh' executed!

• exit code: 0
• output:

Submitted batch job 7392

Test results coming soon (I hope)...

- notification for comment with ID 981834022 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

[boegelbot]

Test report by @boegelbot
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
cnx1 - Linux rocky linux 8.4, x86_64, Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz (haswell), Python 3.6.8
See https://gist.github.com/6217fb2836131925005550af9a2dccdd for a full test report.

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
node3505.doduo.os - Linux RHEL 8.2, x86_64, AMD EPYC 7552 48-Core Processor (zen2), Python 3.6.8
See https://gist.github.com/c1f5704e3926e68a8a8d7e817f6aff43 for a full test report.

[boegel]

Test report by @boegel
FAILED
Build succeeded for 0 out of 1 (1 easyconfigs in total)
node3112.skitty.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz (skylake_avx512), Python 3.6.8
See https://gist.github.com/4defbecaf0e907f8ae053fdcc5926354 for a full test report.

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
node2621.swalot.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) CPU E5-2660 v3 @ 2.60GHz (haswell), Python 3.6.8
See https://gist.github.com/f816c43472899874b726b4bb079bb66e for a full test report.

[migueldiascosta]

so, a couple of tests fail on skylake? or was there another reason for that failed test report?

[boegel]

Test report by @boegel
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
node3136.skitty.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz (skylake_avx512), Python 3.6.8
See https://gist.github.com/e0d7721680046c6d8d83ddc15cc104de for a full test report.

[migueldiascosta]

I guess that answers my question, even though I'm sure there's no causal relation there :)

what happened in the failed test then?

[SebastianAchilles]

Test report by @SebastianAchilles
SUCCESS
Build succeeded for 1 out of 1 (1 easyconfigs in total)
jsfc01.int.jusuf.sebastian.cluster - Linux rocky linux 8.4, x86_64, AMD EPYC 7742 64-Core Processor (zen2), 1 x NVIDIA GRID V100-4C, 460.73.01, Python 3.6.8
See https://gist.github.com/671c3911fa8da49d1408b58e98384df0 for a full test report.

[boegel]

I guess that answers my question, even though I'm sure there's no causal relation there :)

what happened in the failed test then?

I'm not sure, I didn't make any changes, I just resubmitted the test with the same resources requested.
It's not the first time that the PyTorch test suite doesn't seem 100% stable...

For future reference (I don't have the details on how these tests failed anymore)

distributed/algorithms/test_join failed!
distributed/rpc/test_tensorpipe_agent failed!

@migueldiascosta Since the tests have passed fine on a variety of other systems, I think we can assume it was some fluke, and go ahead and merge?

[migueldiascosta]

lgtm

[migueldiascosta]

Going in, thanks @boegel!