Support nn_module_stack in torch.export(strict=False)#115454
Support nn_module_stack in torch.export(strict=False)#115454tugsbayasgalan wants to merge 14 commits intogh/tugsbayasgalan/178/basefrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/115454
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Merge Blocking SEVsThere is 1 active merge blocking SEVs. Please view them below:
If you must merge, use ✅ You can merge normally! (2 Unrelated Failures)As of commit 874ec75 with merge base 1474eb5 ( UNSTABLE - The following jobs failed but were likely due to flakiness present on trunk and has been marked as unstable:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
[ghstack-poisoned]
[ghstack-poisoned]
| return super().getattr(attr, attr_val, parameter_proxy_cache) | ||
| return self.proxy_type(attr_val, attr) | ||
|
|
||
| def call_module(self, m, forward, args, kwargs): |
There was a problem hiding this comment.
This excludes the preserve_signature handling of the version in aot_export. We need this logic, right? Otherwise preserve_signature does not function correctly for non-strict mode?
I'm okay with taking this as a follow-up, but let's make sure to track it
There was a problem hiding this comment.
Yep will do that as follow up.
[ghstack-poisoned]
| torch.set_autocast_cache_enabled(old_value) | ||
|
|
||
|
|
||
| class ModuleStackTracer(PythonKeyTracer): |
There was a problem hiding this comment.
cc @zou3519 (Ed's out - anyone else you think should review this?)
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
@pytorchbot revert -m "Breaking internal tests recycle_bin_citadel and executorch, check internal diff to see more details" -c nosignal |
|
@pytorchbot successfully started a revert job. Check the current status here. |
Reverting PR 115454 failedReason: Command Details for Dev Infra teamRaised by workflow job |
Pull Request resolved: #115188 Approved by: https://github.com/bdhirsh
This reverts commit a267d67. Reverted #115188 on behalf of https://github.com/jeanschmidt due to sadly, it is required to revert this commit in order to revert #115454 ([comment](#115188 (comment)))
|
@pytorchbot revert -m "Breaking internal tests recycle_bin_citadel and executorch, check internal diff to see more details" -c nosignal |
|
@pytorchbot successfully started a revert job. Check the current status here. |
|
@tugsbayasgalan your PR has been successfully reverted. |
This reverts commit 6730b5b. Reverted #115454 on behalf of https://github.com/jeanschmidt due to Breaking internal tests recycle_bin_citadel and executorch, check internal diff to see more details ([comment](#115454 (comment)))
|
cc: @huydhn @mergennachin do you know why wasn't it caught in OSS, as we run executorch builds there now, don't we? |
Pull Request resolved: pytorch#115454 Approved by: https://github.com/suo, https://github.com/bdhirsh
Summary: X-link: pytorch/pytorch#115454 Approved by: https://github.com/suo, https://github.com/bdhirsh Differential Revision: D52339616 Pulled By: tugsbayasgalan
The build was green https://github.com/pytorch/pytorch/actions/runs/7269905380/job/19808210062 on OSS and we pre-build ExecuTorch in Docker. However, it was built with against a 2-week old commit from what I see https://github.com/pytorch/pytorch/blob/6730b5bcb41e0519572759d9ad9852a113d0a7e4/.ci/docker/ci_commit_pins/executorch.txt. This could explain why OSS CI missed it. ExecuTorch on fbcode, on the other hand, used a newer commit for sure. This PR #115599 was supposed to land ExecuTorch latest main commit, but I think GitHub had a hiccup here and It looks ok now in the recent updates:
Having a way to notify oncall about repeated failed attempts to merge an auto commit hash update might be the answer here (stale commit) |
Stack from ghstack (oldest at bottom):
cc @avikchaudhuri @gmagogsfm @zhxchen17 @angelayi @suo @ydwu4