[sparse][semi-structured][inductor] meta registrations for _cslt_sparse_mm + additional stride checking in test. #114685
Conversation
_cslt_sparse_mm + additional stride checking in test. Summary: This PR adds in meta registrations for _cslt_sparse_mm. Based on the work @drisspg did in #114370. Additionally, it updates the tests by checking that the strides of the spare result and the result returned by sparse+compile are the same, to avoid errors like those found in #114477. Test Plan: ``` python test/test_sparse_semi_structred -k compile_cusparselt python test/test_sparse_semi_structred -k compile_cutlass ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/114685
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 20461ef with merge base cef79c0 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
jcaip
changed the title
torch/_meta_registrations.py
Outdated
| out_dtype: Optional[torch.dtype] = None, | ||
| transpose_result: bool = False, | ||
| ): | ||
| assert len(dense_B.shape) == 2, "cslt_sparse_mm only supports 2d inputs" |
There was a problem hiding this comment.
Maybe replicate the two other checks found in the C++ func
There was a problem hiding this comment.
I think it looks good modulo Daniels comments. If you wanted to get much more rigours checking for you, you can add a opinfo in common_method_invocations for the aten op and this will trigger many tests, one being the fake_tensor checking which ensures the meta registration is correct on your sample inputs
… _cslt_sparse_mm + additional stride checking in test." _cslt_sparse_mm + additional stride checking in test. Summary: This PR adds in meta registrations for _cslt_sparse_mm. Based on the work drisspg did in #114370. Additionally, it updates the tests by checking that the strides of the spare result and the result returned by sparse+compile are the same, to avoid errors like those found in #114477. Test Plan: ``` python test/test_sparse_semi_structred -k compile_cusparselt python test/test_sparse_semi_structred -k compile_cutlass ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
… _cslt_sparse_mm + additional stride checking in test." _cslt_sparse_mm + additional stride checking in test. Summary: This PR adds in meta registrations for _cslt_sparse_mm. Based on the work drisspg did in #114370. Additionally, it updates the tests by checking that the strides of the spare result and the result returned by sparse+compile are the same, to avoid errors like those found in #114477. Test Plan: ``` python test/test_sparse_semi_structred -k compile_cusparselt python test/test_sparse_semi_structred -k compile_cutlass ``` Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
…se_mm + additional stride checking in test. Summary: This PR adds in meta registrations for _cslt_sparse_mm. Based on the work drisspg did in #114370. Additionally, it updates the tests by checking that the strides of the spare result and the result returned by sparse+compile are the same, to avoid errors like those found in #114477. Test Plan: ``` python test/test_sparse_semi_structred -k compile_cusparselt python test/test_sparse_semi_structred -k compile_cutlass ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: cc03188 Pull Request resolved: #114685
|
@pytorchbot merge |
Merge failedReason: Approval needed from one of the following: |
|
@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 |
jcaip
added
release notes: sparse
|
@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 |
5 participants
Stack from ghstack (oldest at bottom):
_cslt_sparse_mm + additional stride checking in test.
Summary:
This PR adds in meta registrations for _cslt_sparse_mm.
Based on the work @drisspg did
in #114370.
Additionally, it updates the tests by checking that the strides of the
spare result and the result returned by sparse+compile are the same, to
avoid errors like those found in
#114477.
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags: