[pt2] add meta for argsort.stable, use sort samples in OpInfo
#106025
Conversation
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/106025
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 4264518: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
From the list in #105105 |
| @@ -3739,6 +3739,7 @@ def test_vmap_exhaustive(self, device, dtype, op): | |||
| # but found at least two devices, cuda:0 and cpu! | |||
| xfail('ge', device_type='cuda'), | |||
| xfail('_upsample_bilinear2d_aa'), | |||
| xfail('argsort'), # aten::argsort.stable hit the vmap fallback which is currently disabled | |||
There was a problem hiding this comment.
OpInfo now tests with the stable kwarg, so this got uncovered.
| @@ -18269,7 +18265,7 @@ def reference_flatten(input, start_dim=0, end_dim=-1): | |||
| "argsort", | |||
| dtypes=all_types_and(torch.bool, torch.float16, torch.bfloat16), | |||
| dtypesIfCUDA=all_types_and(torch.float16, torch.bfloat16), | |||
| sample_inputs_func=sample_inputs_argsort, | |||
| sample_inputs_func=sample_inputs_sort, | |||
There was a problem hiding this comment.
argsort is the second value returned by torch.sort(). Same API as sort, not sure why samples weren't reused before.
torch.argsort(input, dim=- 1, descending=False, stable=False) → Tensor
torch.sort(input, dim=- 1, descending=False, stable=False, *, out=None)
Note: on the aten side, stable has no default:
func: argsort.stable(Tensor self, *, bool stable, int dim=-1, bool descending=False) -> Tensor
nkaretnikov
requested review from
zou3519,
Chillee,
kshitij12345,
mruberry and
ngimel
as code owners
|
@pytorchbot rebase |
|
@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here |
… `OpInfo`" [ghstack-poisoned]
|
Successfully rebased |
ghstack-source-id: b914535c290c967fe81bb7bf111858bb54d193c3 Pull Request resolved: #106025
|
@pytorchbot merge |
Merge failedReason: This PR needs a If not, please add the To add a label, you can comment to pytorchbot, for example For more information, see Details for Dev Infra teamRaised by workflow job |
|
@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 |
|
The merge job was canceled. If you believe this is a mistake,then you can re trigger it through pytorch-bot. |
|
@pytorchbot merge -f "looks ok" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Stack from ghstack (oldest at bottom):
argsort.stable, usesortsamples inOpInfo#106025