Add onednn quant backend #74137
Closed
Add onednn quant backend #74137
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
…e=True` in qconfig for unit test TestQuantizedOps.test_custom_module_multi_head_attention. Skip unsupported tests (output padding for deconv)
…ackend only supports symmetric quantization of weight
…nednn_quant_backend
…nednn_quant_backend
…nednn_quant_backend
CI Flow Status⚛️ CI FlowRuleset - Version:
|
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 354aa66 (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
jerryzh168
approved these changes
Mar 12, 2022
Contributor
jerryzh168
left a comment
jerryzh168
left a comment
There was a problem hiding this comment.
thanks! looks like previous error was due to a land time merge conflict, it is pretty rare.
Contributor
|
@jerryzh168 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
facebook-github-bot
pushed a commit
that referenced
this pull request
Mar 15, 2022
Contributor
|
Hey @Xia-Weiwen. |
Collaborator
Author
|
Hi @jerryzh168. The bot asks me to add labels to this PR. Is it a must? Looks like I don't have the permission to add labels. |
Contributor
|
oh, this is for release notes, I see, I can add this |
Contributor
|
do we want to advertise this to user now? or maybe announce it later? |
Collaborator
Author
Hi @jerryzh168. Shall we have a discussion on slack? |
vkuzo
added a commit
that referenced
this pull request
Jun 16, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Test plan: ``` python test/test_quantization.py -k Fx ``` [ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Jun 16, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Test plan: ``` python test/test_quantization.py -k Fx ``` ghstack-source-id: 35ca5c3 Pull Request resolved: #79718
vkuzo
added a commit
that referenced
this pull request
Jun 16, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Test plan: ``` python test/test_quantization.py -k Fx ``` [ghstack-poisoned]
vkuzo
added a commit
that referenced
this pull request
Jun 16, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Test plan: ``` python test/test_quantization.py -k Fx ``` ghstack-source-id: acf8585 Pull Request resolved: #79718
pytorchmergebot
pushed a commit
that referenced
this pull request
Jun 16, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Test plan: ``` python test/test_quantization.py -k Fx ``` Pull Request resolved: #79718 Approved by: https://github.com/jerryzh168
facebook-github-bot
pushed a commit
that referenced
this pull request
Jun 20, 2022
Summary: In #74137, the MKLDNN quantized backend was added to PyTorch. Sometime in the past couple of days, MKLDNN got enabled on my Mac OS machine. This uncovered issues in FX graph mode quantization testing, as we were only testing for fbgemm and qnnpack, and some of the tests that were assuming fbgemm started silently going through the MKLDNN path. Since the requirements for MKLDNN are different, the tests started to fail. This PR unbreaks the minimal amount of tests to get a clean test run on my machine. In the future, it would be great to add testing for MKLDNN specifically, and also audit all of the current quantization tests which are assuming fbgemm to set the backend properly. Pull Request resolved: #79718 Approved by: https://github.com/jerryzh168 Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/91c5fc323b933516a46add626cfa402f73d7e474 Test plan from GitHub: ``` python test/test_quantization.py -k Fx ``` Reviewed By: malfet Differential Revision: D37242148 fbshipit-source-id: e1bb95a20b2cc45f1b261f1512f7e24e5bc24640
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
Resolve the conflicts in #69820
@jerryzh168 Please review. Thanks.