[JIT] OpInfo tests for nvfuser #71299
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
CI Flow Status⚛️ CI FlowRuleset - Version:
You can add a comment to the PR and tag @pytorchbot with the following commands: # ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun
# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slowFor more information, please take a look at the CI Flow Wiki. |
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 2a08ab2 (more details on the Dr. CI page):
🕵️♀️ 2 failures not recognized by patterns:The following CI failures may be due to changes from the PR
This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
Contributor
Author
|
current test failures: nvfuser-opinfo.txt ignore the following op failures (which I've disabled now, since they fail on the jit variant consistency tests as well):
|
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
eellison
reviewed
Jan 14, 2022
|
|
||
| _tracing_ops = partial(ops, dtypes=OpDTypes.supported, | ||
| allowed_dtypes=(torch.float, torch.cfloat)) | ||
|
|
Contributor
There was a problem hiding this comment.
is there a reason we're restricting to float and cfloat ?
Contributor
Author
There was a problem hiding this comment.
copied this from the variant_consistency tests, where
# variant testing is only done with torch.float and torch.cfloat to avoid
# excessive test times and maximize signal to noise ratioWhat are your thoughts here, should we expand this to all dtypes?
Contributor
There was a problem hiding this comment.
there's a slow test that runs nightly, we could at least run it then (and initially, to flush out issue)
Contributor
There was a problem hiding this comment.
ctrl-f SLOW_TEST or something and you'll find it
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
davidberard98
changed the title
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Differential Revision: [D33595299](https://our.internmc.facebook.com/intern/diff/D33595299) [ghstack-poisoned]
Contributor
Author
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
eellison
approved these changes
Apr 1, 2022
Contributor
eellison
left a comment
eellison
left a comment
There was a problem hiding this comment.
😍 😍 😍 Should we file issues for the failing tests ?
| # https://github.com/pytorch/pytorch/issues/71784 | ||
| DecorateInfo(unittest.skip('Skipped!'), 'TestNNCOpInfo', 'test_nnc_correctness', | ||
| device_type='cpu', dtypes=(torch.float16,)), | ||
| DecorateInfo(unittest.skip('Skipped!'), 'TestCudaFuserOpInfo', 'test_nvfuser_correctness', dtypes=(torch.float16,)), |
Contributor
There was a problem hiding this comment.
Cool! should we file issues for the failing tests ?
Contributor
Author
There was a problem hiding this comment.
I think most of them either have an issue filed or are expected to fail
e.g. #71784 for this one
Collaborator
There was a problem hiding this comment.
Let me jump on the failing tests!
Contributor
Author
There was a problem hiding this comment.
@jjsjann123 fyi I think #71784 might be expected
And list of tests that need fixes is in #75029 (also see this board: https://github.com/pytorch/pytorch/projects/30)
facebook-github-bot
pushed a commit
that referenced
this pull request
Apr 1, 2022
Summary: Pull Request resolved: #71299 These tests verify that for the same inputs, the eager version of an op and a traced, fused version of the op return the same output. Currently the tests don't check whether or not fusion actually occurred. Test Plan: Imported from OSS Reviewed By: eellison Differential Revision: D33595299 Pulled By: davidberard98 fbshipit-source-id: 26fdacf44941808c134953e7a883a02d13a43f19
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack:
These tests verify that for the same inputs, the eager version of an op
and a traced, fused version of the op return the same output.
Currently the tests don't check whether or not fusion actually occurred.
Differential Revision: D33595299