Skip to content

Improved matmul tests#75193

Closed
lezcano wants to merge 8 commits intogh/Lezcano/55/basefrom
gh/Lezcano/55/head
Closed

Improved matmul tests#75193
lezcano wants to merge 8 commits intogh/Lezcano/55/basefrom
gh/Lezcano/55/head

Conversation

@lezcano
Copy link
Collaborator

@lezcano lezcano commented Apr 4, 2022 •
edited
Loading

Stack from ghstack:

Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the out= variant of matmul is broken
when used on 4D tensors and a number of other cases. This hints what would be the amount of people
that use out= variants...

@lezcano
Improved matmul tests
84c81b3 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

facebook-github-bot commented Apr 4, 2022 •
edited
Loading

🔗 Helpful links

💊 CI failures summary and remediations

As of commit 103cc38 (more details on the Dr. CI page):

Expand to see more
  • 1/1 failures introduced in this PR

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages

See GitHub Actions build pull / linux-xenial-py3.7-gcc5.4 / test (backwards_compat, 1, 1, linux.2xlarge) (1/1)

Step: "Test" (full log | diagnosis details | 🔁 rerun)

2022-04-21T13:27:11.3842714Z The PR is introduc...m to confirm whether this change is wanted or not. 
2022-04-21T13:27:11.3827750Z processing existing schema:  text(__torch__.torch.classes.profiling.SourceRef _0) -> (str _0)
2022-04-21T13:27:11.3829021Z processing existing schema:  count(__torch__.torch.classes.profiling.InstructionStats _0) -> (int _0)
2022-04-21T13:27:11.3831015Z processing existing schema:  duration_ns(__torch__.torch.classes.profiling.InstructionStats _0) -> (int _0)
2022-04-21T13:27:11.3832327Z processing existing schema:  source(__torch__.torch.classes.profiling.SourceStats _0) -> (__torch__.torch.classes.profiling.SourceRef _0)
2022-04-21T13:27:11.3834970Z processing existing schema:  line_map(__torch__.torch.classes.profiling.SourceStats _0) -> (Dict(int, __torch__.torch.classes.profiling.InstructionStats) _0)
2022-04-21T13:27:11.3835694Z processing existing schema:  __init__(__torch__.torch.classes.profiling._ScriptProfile _0) -> (NoneType _0)
2022-04-21T13:27:11.3837330Z processing existing schema:  enable(__torch__.torch.classes.profiling._ScriptProfile _0) -> (NoneType _0)
2022-04-21T13:27:11.3838584Z processing existing schema:  disable(__torch__.torch.classes.profiling._ScriptProfile _0) -> (NoneType _0)
2022-04-21T13:27:11.3841200Z processing existing schema:  _dump_stats(__torch__.torch.classes.profiling._ScriptProfile _0) -> (__torch__.torch.classes.profiling.SourceStats[] _0)
2022-04-21T13:27:11.3842330Z processing existing schema:  __init__(__torch__.torch.classes.dist_rpc.WorkerInfo _0, str _1, int _2) -> (NoneType _0)
2022-04-21T13:27:11.3842714Z The PR is introducing backward incompatible changes to the operator library. Please contact PyTorch team to confirm whether this change is wanted or not. 
2022-04-21T13:27:11.3842725Z 
2022-04-21T13:27:11.3842828Z Broken ops: [
2022-04-21T13:27:11.3843219Z 	aten::expand.SymInt(Tensor(a) self, SymInt[] size, *, bool implicit=False) -> (Tensor(a))
2022-04-21T13:27:11.3843501Z 	aten::chalf(Tensor self, *, int? memory_format=None) -> (Tensor)
2022-04-21T13:27:11.3843819Z 	aten::_masked_softmax(Tensor self, Tensor mask, int? dim=None) -> (Tensor)
2022-04-21T13:27:11.3844219Z 	aten::_masked_softmax_backward(Tensor grad_output, Tensor output, Tensor mask, int? dim=None) -> (Tensor)
2022-04-21T13:27:11.3844572Z 	aten::expand_copy.SymInt(Tensor self, SymInt[] size, *, bool implicit=False) -> (Tensor)
2022-04-21T13:27:11.3844664Z ]
2022-04-21T13:27:11.5181849Z + cleanup
2022-04-21T13:27:11.5182336Z + retcode=1
This comment was automatically generated by Dr. CI (expand for details).

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

This was referenced Apr 4, 2022
Closed
Closed
Closed
Closed
Closed
@lezcano lezcano requested review from ngimel and removed request for IvanYashchuk and nikitaved April 4, 2022 16:23
@lezcano lezcano added module: linear algebra Issues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmul module: tests Issues related to tests (not the torch.testing module) topic: not user facing topic category labels Apr 4, 2022
@lezcano
Update on "Improved matmul tests"
45d23db 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano lezcano mentioned this pull request Apr 4, 2022
Closed
@lezcano
Update on "Improved matmul tests"
6101bc0 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano lezcano mentioned this pull request Apr 4, 2022
Closed
lezcano added 2 commits April 4, 2022 22:01
@lezcano
Update on "Improved matmul tests"
0684364 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano
Update on "Improved matmul tests"
99fa829 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano lezcano mentioned this pull request Apr 5, 2022
Closed
lezcano added 3 commits April 6, 2022 14:22
@lezcano
Update on "Improved matmul tests"
f1fbc5e 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano
Update on "Improved matmul tests"
c214c94 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
@lezcano
Update on "Improved matmul tests"
103cc38 
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors and a number of other cases. This hints what would be the amount of people
that use out= variants...

[ghstack-poisoned]
ngimel
ngimel approved these changes May 3, 2022
View reviewed changes
Copy link
Collaborator

@ngimel ngimel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How long do added tests take?

@lezcano
Copy link
Collaborator Author

lezcano commented May 3, 2022

17 seconds

@ngimel
Copy link
Collaborator

ngimel commented May 3, 2022

@pytorchbot merge this

facebook-github-bot pushed a commit that referenced this pull request May 4, 2022
@lezcano @facebook-github-bot
Improved matmul tests (#75193)
2535b77 
Summary:
Let's make sure we don't break anything in the next PRs of the stack.
Also some comprehensive testing of matmul on CPU and CUDA was long due.

Running this tests we see that the `out=` variant of matmul is broken
when used on 4D tensors. This hints what would be the amount of people
that use out= variants...

Pull Request resolved: #75193

Approved by: https://github.com/ngimel

Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/9e34a8241b5f18fd47a64349d955c90c8637dc14

Reviewed By: malfet

Differential Revision: D36134075

fbshipit-source-id: a9c5b1a5c391f9e8a1bde42c7e8795e745e1d007
@facebook-github-bot facebook-github-bot deleted the gh/Lezcano/55/head branch May 7, 2022 14:17
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ngimel ngimel ngimel approved these changes

Assignees

No one assigned

Labels

cla signed module: linear algebra Issues related to specialized linear algebra operations in PyTorch; includes matrix multiply matmul module: tests Issues related to tests (not the torch.testing module) open source topic: not user facing topic category

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@lezcano @facebook-github-bot @ngimel @pytorchbot