Move Concat Linear out of Optimize Numerics #67196
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
[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. |
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit ad0827e (more details on the Dr. CI page):
2 failures not recognized by patterns:
This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
facebook-github-bot
added
the
oncall: jit
eellison
reviewed
Oct 25, 2021
There was a problem hiding this comment.
can you just remove https://github.com/pytorch/pytorch/blob/master/torch/jit/_freeze.py#L162 and have it call into OptimizeFrozenGraph ? originally there was some thought of python hackability etc, but we should just remove the duplication.
can you also hook up the pretranspose pass?
[ghstack-poisoned]
eellison
approved these changes
Oct 26, 2021
eellison
requested changes
Oct 26, 2021
eellison
reviewed
Oct 26, 2021
There was a problem hiding this comment.
we're probably going to need to remove the MKLDNN Linear tests, that's fine. can we remove https://github.com/pytorch/pytorch/blob/master/torch/csrc/jit/passes/frozen_ops_to_mkldnn.cpp#L939 the linear here and say that we get more speedup by pretransposing linear layers than running them in mkldnn
[ghstack-poisoned]
eellison
reviewed
Oct 27, 2021
| test_unsupported(nn.Sequential(lin, Add(torch.tensor([20]))), ['1']) | ||
|
|
||
| @unittest.skipIf(not torch._C.has_mkldnn, "MKL-DNN build is disabled") | ||
| def test_mkldnn_fuser_broadcasting(self): |
| self.assertEqual(mod_to_device(*test_vals_to_device), script_mod(*test_vals_to_device)) | ||
|
|
||
| @unittest.skipIf(not torch._C.has_mkldnn, "MKL-DNN build is disabled") | ||
| def test_collapse_adjacent_conversions(self): |
| self.assertEqual(aten_op(x, inplace=False), m(x).to_dense()) | ||
|
|
||
| @unittest.skipIf(not torch._C.has_mkldnn, "MKL-DNN build is disabled") | ||
| def test_scalar_mul(self): |
[ghstack-poisoned]
|
hey - we have an internal user who's optimization is blocked by this. could we try to land it ? |
[ghstack-poisoned]
[ghstack-poisoned]
Gamrix
commented
Nov 3, 2021
| The current set of optimizations is: | ||
| The current set of optimizations includes: | ||
| - Dropout Removal | ||
| - Pretranspose Linear Layers |
[ghstack-poisoned]
[ghstack-poisoned]
Closed
|
@Gamrix has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
|
This pull request has been reverted by 3d4a6ff. To re-land this change, follow these steps. |
|
This pull request has been reverted by 3d4a6ff. To re-land this change, follow these steps. |
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:
Differential Revision: D32154788