-
Notifications
You must be signed in to change notification settings - Fork 21.4k
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
Dispatch take_along_axis to gather #107711
Conversation
Gather does the same thing, but it's much better supported in the `torch.compile` stack [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/107711
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit e8d25cd with merge base ba5eeed (): This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Gather does the same thing, but it's much better supported in the `torch.compile` stack cc mruberry rgommers voznesenskym penguinwu anijain2305 EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng Xia-Weiwen wenzhe-nrv jiayisunx chenyang78 aakhundov [ghstack-poisoned]
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
tests?
Gather does the same thing, but it's much better supported in the `torch.compile` stack cc mruberry rgommers voznesenskym penguinwu anijain2305 EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng Xia-Weiwen wenzhe-nrv jiayisunx chenyang78 aakhundov [ghstack-poisoned]
yield SampleInput( | ||
make_arg(), | ||
0, | ||
torch.tensor([0], dtype=torch.int64, device=device)) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Repeated sample.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@lezcano one of these is torch.tensor([0])
(i.e. 1d) and the other is torch.tensor(0)
(i.e. 0d)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah, right, sorry for that. Reverting at #107776
@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 |
Merge failedReason: 1 mandatory check(s) failed. The first few are: Dig deeper by viewing the failures on hud |
Gather does the same thing, but it's much better supported in the `torch.compile` stack cc mruberry rgommers voznesenskym penguinwu anijain2305 EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng Xia-Weiwen wenzhe-nrv jiayisunx chenyang78 aakhundov [ghstack-poisoned]
@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 |
This seems to fail on main locally:
Reverting the change from |
It does! Let me put up a forward fix. This means that not all the tests in |
Addresses #107711 (comment) I'll have someone submit a decomposition of `take_along_dim` next week. [ghstack-poisoned]
Addresses #107711 (comment) I'll have someone submit a decomposition of `take_along_dim` next week. ghstack-source-id: f59e75463653db9086a519c46165afe1ce926136 Pull Request resolved: #107875
Stack from ghstack (oldest at bottom):
Gather does the same thing, but it's much better supported in the
torch.compile
stackcc @mruberry @rgommers @voznesenskym @penguinwu @anijain2305 @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @Xia-Weiwen @wenzhe-nrv @jiayisunx @chenyang78 @aakhundov