{tools}[foss/2024a] PyTorch v2.6.0, parameterized v0.9.0, optree v0.14.1, ...#22824
Merged
lexming merged 12 commits intoeasybuilders:developfrom Jul 30, 2025
Merged
Conversation
…GCCcore-13.3.0.eb, optree-0.14.1-GCCcore-13.3.0.eb, pytest-rerunfailures-15.0-GCCcore-13.3.0.eb, tlparse-0.3.37-GCCcore-13.3.0.eb, pytest-subtests-0.13.1-GCCcore-13.3.0.eb, pytest-shard-0.1.2-GCCcore-13.3.0.eb and patches: PyTorch-2.6.0_add-checkfunctionexists-include.patch, PyTorch-2.6.0_avoid_caffe2_test_cpp_jit.patch, PyTorch-2.6.0_disable_DataType_dependent_test_if_tensorboard_is_not_available.patch, PyTorch-2.6.0_disable-gcc12-warnings.patch, PyTorch-2.6.0_disable_tests_which_need_network_download.patch, PyTorch-2.6.0_fix-accuracy-issues-in-linalg_solve.patch, PyTorch-2.6.0_fix-distributed-tests-without-gpus.patch, PyTorch-2.6.0_fix-edge-case-causing-test_trigger_bisect_on_error-failure.patch, PyTorch-2.6.0_fix-ExcTests.test_trigger_on_error.patch, PyTorch-2.6.0_fix-flaky-test_aot_export_with_torch_cond.patch, PyTorch-2.6.0_fix-inductor-device-interface.patch, PyTorch-2.6.0_fix-server-in-test_control_plane.patch, PyTorch-2.6.0_fix-skip-decorators.patch, PyTorch-2.6.0_fix-sympy-1.13-compat.patch, PyTorch-2.6.0_fix-test_autograd_cpp_node_saved_float.patch, PyTorch-2.6.0_fix-test_linear_with_embedding.patch, PyTorch-2.6.0_fix-test_linear_with_in_out_buffer-without-mkl.patch, PyTorch-2.6.0_fix-test_public_bindings.patch, PyTorch-2.6.0_fix-test_unbacked_bindings_for_divisible_u_symint.patch, PyTorch-2.6.0_fix-vsx-vector-shift-functions.patch, PyTorch-2.6.0_fix-xnnpack-float16-convert.patch, PyTorch-2.6.0_increase-tolerance-test_aotdispatch-matmul.patch, PyTorch-2.6.0_increase-tolerance-test_quick-baddbmm.patch, PyTorch-2.6.0_increase-tolerance-test_vmap_autograd_grad.patch, PyTorch-2.6.0_no-cuda-stubs-rpath.patch, PyTorch-2.6.0_remove-test_slice_with_floordiv.patch, PyTorch-2.6.0_skip-diff-test-on-ppc.patch, PyTorch-2.6.0_skip-test_checkpoint_wrapper_parity-on-cpu.patch, PyTorch-2.6.0_skip-test_init_from_local_shards.patch, PyTorch-2.6.0_skip-test_jvp_linalg_det_singular.patch, PyTorch-2.6.0_skip-test-requiring-MKL.patch, PyTorch-2.6.0_skip-test_segfault.patch, PyTorch-2.6.0_skip-tests-without-fbgemm.patch
Updated software
|
1 task
Member
|
@boegelbot please test @ jsc-zen3 |
Collaborator
|
@boegel: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 2858907011 processed Message to humans: this is just bookkeeping information for me, |
Contributor
Author
|
There might be another problem I haven't seen before: |
Collaborator
|
Test report by @boegelbot |
Contributor
Author
|
Test report by @Flamefire |
Contributor
|
Test report by @lexming |
Contributor
|
Test report by @lexming |
Contributor
Author
|
@lexming I see an unusually large number of failed tests. Do you have any details on that? As for the blocking error:
I'll look into the easyblock and relax the check if I cannot fix the issue. This test is a kind of sanity-check whether the results collected look complete and reasonable. Hence the hard failure to force looking into it instead of silently missing e.g. a large number of test results. |
Contributor
|
I got the following error: Any ideas about what the issue is? |
Contributor
Author
|
Looks similar to the results of @lexming |
Contributor
|
Here's the log file: easybuild-PyTorch-2.6.0-20250516.130225.QgNtC.log.gz |
boegel
changed the title
Contributor
Author
|
@tardigradus In your log I see: This is a pytest bug for which I added a patch to our existing module: #22602 @lexming Probably the same for you looking at the similar failures
This should be fixed with easybuilders/easybuild-easyblocks#3723 |
Contributor
|
I have reinstalled pytest as suggested, but now fails with What am I doing wrong? |
Contributor
Author
You need to start of the develop branch. IIRC the PR containing that patch isn't in a release yet. Or in a release newer than what you are using. |
Contributor
Thanks for the hints. I have tried downloading the patch, but then I get errors about other patch files being missing. So perhaps I should "start of(f?) the develop branch", but what exactly does that mean? Is it covered in the documentation? If so, could you please let me know which part? |
Contributor
Author
|
Basically: All PRs might rely on files/changes not in any release yet. So if you don't use "something" with those changes you'll see issues. Otherwise install the current develop branch: https://docs.easybuild.io/installation-alternative/?h=install+easybuild+develop#installation-of-latest-development-version The next best option that might work is run easybuild with In both cases: New easyconfigs might require new easyblocks so this might fail anyway so better install all 3 repos (easyconfigs, easyblocks, framework) from the develop branch as described in the doc above. |
Contributor
|
Thanks for the clarification. As the notes from the 2025.05.21 conference call indicate that the next release is nigh, I'll probably just wait for that and see how I get on. |
Contributor
|
I have upgraded to EB 5.1.0. However, PyTorch 2.6.0 has a strict dependency on Would a solution be to create |
Contributor
|
patch file PyTorch-2.6.0_fix-cpuinfo-bug-with-smt.patch is missing. |
Contributor
Author
|
Test report by @Flamefire |
Contributor
Author
|
Test report by @Flamefire |
Contributor
Author
|
Test report by @Flamefire |
smoors
reviewed
Jul 4, 2025
Contributor
Author
|
Test report by @Flamefire |
Contributor
|
This build fails if |
1 task
Contributor
Author
No we don't. Why/How does it fail? It works for me (see above test reports) |
Contributor
|
@boegelbot please test @ jsc-zen3 |
Collaborator
|
@lexming: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 3125910886 processed Message to humans: this is just bookkeeping information for me, |
Contributor
|
Test report by @lexming |
Collaborator
|
Test report by @boegelbot |
Contributor
|
Merging, thanks for the hard work @Flamefire ! |
Contributor
Author
|
easybuilders/easybuild-easyblocks#3803 is also relevant which avoids unnecessary failures. I'm updating that while testing PT 2.7 |
4 tasks
Contributor
attaching the eb logs, we seem to be way over the 16 baseline for errors allowed |
Contributor
Author
|
It fails first because it misses a failed test. That should be inside One issue needs a rebuild of a dependent module: #22602 Then there are multiple What I'd try (on this machine) is running the test with the pypi package:
|
Member
|
@arielzn It's hard to keep track of discussions like this in merged PRs, can we open an issue on this instead? |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
9 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
(created using
eb --new-pr)