Unify quantized conv and linear tests #26992
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
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
pytorchbot
added
oncall: quantization
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Sep 27, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 82b4f7e Pull Request resolved: #26992
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Sep 28, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 33db9b2 Pull Request resolved: #26992
dzhulgakov
reviewed
Sep 28, 2019
dzhulgakov
reviewed
Sep 28, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK python test/test_quantized.py TestQuantizedConv ---------------------------------------------------------------------- Ran 2 tests in 0.108s OK Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Sep 30, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. For op level as well as module level tests. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 3810440 Pull Request resolved: #26992
dskhudia
reviewed
Sep 30, 2019
Summary:
Run the same test for FBGEMM and QNNPACK backends.
Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines)
Test Plan:
python test/test_quantized.py TestQuantizedLinear
python test/test_quantized.py TestQuantizedConv
python test/test_quantized_models.py
python test/test_quantized_nn_mods.py
Reviewers:
Subscribers:
Tasks:
Tags:
[ghstack-poisoned]
Summary:
Run the same test for FBGEMM and QNNPACK backends.
Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines)
Test Plan:
python test/test_quantized.py TestQuantizedLinear
python test/test_quantized.py TestQuantizedConv
python test/test_quantized_models.py
python test/test_quantized_nn_mods.py
Reviewers:
Subscribers:
Tasks:
Tags:
[ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Oct 1, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. For op level as well as module level tests. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 4435ff7 Pull Request resolved: #26992
Summary:
Run the same test for FBGEMM and QNNPACK backends.
Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines)
Test Plan:
python test/test_quantized.py TestQuantizedLinear
python test/test_quantized.py TestQuantizedConv
python test/test_quantized_models.py
python test/test_quantized_nn_mods.py
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D17689171](https://our.internmc.facebook.com/intern/diff/D17689171)
[ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Oct 1, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. For op level as well as module level tests. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 4077938 Pull Request resolved: #26992
jerryzh168
reviewed
Oct 1, 2019
jerryzh168
reviewed
Oct 1, 2019
| # Setting target SQNR to be 30 dB so that relative error is 1e-3 below the desired | ||
| # output | ||
| self.assertGreater(SQNRdB, 30, msg='Quantized model numerics diverge from float, expect SQNR > 30 dB') | ||
| for qengine in ["fbgemm", "qnnpack"]: |
There was a problem hiding this comment.
Also, this is an integration test, I think you should put this in torch.quantization, but this can be done in a separate PR
jerryzh168
reviewed
Oct 1, 2019
| # output | ||
| self.assertGreater(SQNRdB, 30, msg='Quantized model numerics diverge from float, expect SQNR > 30 dB') | ||
| for qengine in ["fbgemm", "qnnpack"]: | ||
| if qengine not in torch.backends.quantized.supported_engines: |
There was a problem hiding this comment.
Also, shouldn't we use randomized tests to test numerics?
There was a problem hiding this comment.
I haven't modified the test really. Just run the existing test for different qengine.
cc @raghuramank100 for adding randomized tests.
jerryzh168
reviewed
Oct 1, 2019
jerryzh168
approved these changes
Oct 1, 2019
Summary:
Run the same test for FBGEMM and QNNPACK backends.
Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines)
Test Plan:
python test/test_quantized.py TestQuantizedLinear
python test/test_quantized.py TestQuantizedConv
python test/test_quantized_models.py
python test/test_quantized_nn_mods.py
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: [D17689171](https://our.internmc.facebook.com/intern/diff/D17689171)
[ghstack-poisoned]
supriyar
added a commit
that referenced
this pull request
Oct 1, 2019
Summary: Run the same test for FBGEMM and QNNPACK backends. For op level as well as module level tests. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 571a252 Pull Request resolved: #26992
|
This pull request has been merged in b805b5d. |
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26992 Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Imported from OSS Differential Revision: D17689171 fbshipit-source-id: e11c0a5e41f5f4e6836a614a5b61e4db3c5e384b
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26992 Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Imported from OSS Differential Revision: D17689171 fbshipit-source-id: e11c0a5e41f5f4e6836a614a5b61e4db3c5e384b
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26992 Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Imported from OSS Differential Revision: D17689171 fbshipit-source-id: e11c0a5e41f5f4e6836a614a5b61e4db3c5e384b
soumith
pushed a commit
that referenced
this pull request
Oct 7, 2019
Summary: Pull Request resolved: #26992 Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Imported from OSS Differential Revision: D17689171 fbshipit-source-id: e11c0a5e41f5f4e6836a614a5b61e4db3c5e384b
pdlive215
pushed a commit
to pdlive215/pytorch
that referenced
this pull request
Nov 27, 2019
Summary: Pull Request resolved: pytorch#26992 Run the same test for FBGEMM and QNNPACK backends. Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines) Test Plan: python test/test_quantized.py TestQuantizedLinear python test/test_quantized.py TestQuantizedConv python test/test_quantized_models.py python test/test_quantized_nn_mods.py Imported from OSS Differential Revision: D17689171 fbshipit-source-id: e11c0a5e41f5f4e6836a614a5b61e4db3c5e384b
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
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:
Summary:
Run the same test for FBGEMM and QNNPACK backends.
Checks that QNNPACK or FBGEMM are supported before running it (using supported_qengines)
Test Plan:
python test/test_quantized.py TestQuantizedLinear
python test/test_quantized.py TestQuantizedConv
python test/test_quantized_models.py
python test/test_quantized_nn_mods.py
Reviewers:
Subscribers:
Tasks:
Tags:
Differential Revision: D17689171