[quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

andrewor14 · 2023-09-22T20:12:00Z

Stack from ghstack (oldest at bottom):

-> [quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

Summary: Today, we get different batch norm ops depending on
the device the model is placed on at export time. Exporting
model.cpu() gives _native_batch_norm_legit, while exporting
model.cuda() gives cudnn_batch_norm. QAT fusion currently
only supports the former and silently ignores the latter. This
commit fixes this by additionally matching on the latter op
during QAT fusion.

Test Plan:
python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion
python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion

Reviewers: jerryzh168, kimishpatel

Subscribers: jerryzh168, kimishpatel, supriyar

Differential Revision: D49615145

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar [ghstack-poisoned]

pytorch-bot · 2023-09-22T20:12:04Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/109908

📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

Note: Links to docs will display an error until the docs builds have been completed.

✅ You can merge normally! (3 Unrelated Failures)

As of commit f7eca9c with merge base 156aefa ():

UNSTABLE - The following jobs failed but were likely due to flakiness present on trunk and has been marked as unstable:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar [ghstack-poisoned]

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar ghstack-source-id: be4fa0b Pull Request resolved: #109908

jerryzh168

looks good!

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar [ghstack-poisoned]

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar ghstack-source-id: fe5f6bd Pull Request resolved: #109908

andrewor14 · 2023-09-25T22:05:54Z

@andrewor14 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar Differential Revision: [D49615145](https://our.internmc.facebook.com/intern/diff/D49615145) [ghstack-poisoned]

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar ghstack-source-id: 64ec3b1 Pull Request resolved: #109908

andrewor14 · 2023-09-27T20:46:31Z

@andrewor14 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar Differential Revision: [D49615145](https://our.internmc.facebook.com/intern/diff/D49615145) [ghstack-poisoned]

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar ghstack-source-id: 515b641 Pull Request resolved: #109908

andrewor14 · 2023-09-28T18:18:19Z

@andrewor14 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar Differential Revision: [D49615145](https://our.internmc.facebook.com/intern/diff/D49615145) [ghstack-poisoned]

Summary: Today, we get different batch norm ops depending on the device the model is placed on at export time. Exporting `model.cpu()` gives `_native_batch_norm_legit`, while exporting `model.cuda()` gives `cudnn_batch_norm`. QAT fusion currently only supports the former and silently ignores the latter. This commit fixes this by additionally matching on the latter op during QAT fusion. Test Plan: python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_fusion python test/test_quantization.py TestQuantizePT2EQAT.test_qat_conv_bn_relu_fusion Reviewers: jerryzh168, kimishpatel Subscribers: jerryzh168, kimishpatel, supriyar ghstack-source-id: 5806fde Pull Request resolved: #109908

andrewor14 · 2023-10-04T15:04:41Z

@andrewor14 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

andrewor14 · 2023-10-05T01:34:17Z

@pytorchbot merge

pytorchmergebot · 2023-10-05T01:36:17Z

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

Fixes recent broken unit tests caused by PR #109908 because cudnn and miopen have separate batch norm functions. ``` 2023-10-05T09:35:01.6606614Z _______________ TestQuantizePT2EQAT.test_qat_conv_bn_fusion_cuda _______________ 2023-10-05T09:35:01.6606948Z Traceback (most recent call last): 2023-10-05T09:35:01.6607362Z File "/var/lib/jenkins/pytorch/test/quantization/pt2e/test_quantize_pt2e_qat.py", line 323, in test_qat_conv_bn_fusion_cuda 2023-10-05T09:35:01.6607767Z self._verify_symmetric_xnnpack_qat_graph( 2023-10-05T09:35:01.6608217Z File "/var/lib/jenkins/pytorch/test/quantization/pt2e/test_quantize_pt2e_qat.py", line 130, in _verify_symmetric_xnnpack_qat_graph 2023-10-05T09:35:01.6608658Z self._verify_symmetric_xnnpack_qat_graph_helper( 2023-10-05T09:35:01.6609105Z File "/var/lib/jenkins/pytorch/test/quantization/pt2e/test_quantize_pt2e_qat.py", line 173, in _verify_symmetric_xnnpack_qat_graph_helper 2023-10-05T09:35:01.6609623Z m = prepare_qat_pt2e(m, quantizer) 2023-10-05T09:35:01.6610171Z File "/opt/conda/envs/py_3.8/lib/python3.8/site-packages/torch/ao/quantization/quantize_pt2e.py", line 178, in prepare_qat_pt2e 2023-10-05T09:35:01.6610561Z _fuse_conv_bn_qat(model) 2023-10-05T09:35:01.6611072Z File "/opt/conda/envs/py_3.8/lib/python3.8/site-packages/torch/ao/quantization/pt2e/qat_utils.py", line 501, in _fuse_conv_bn_qat 2023-10-05T09:35:01.6611497Z m = _fuse_conv_bn_qat_helper(m, is_cuda=True) 2023-10-05T09:35:01.6612065Z File "/opt/conda/envs/py_3.8/lib/python3.8/site-packages/torch/ao/quantization/pt2e/qat_utils.py", line 575, in _fuse_conv_bn_qat_helper 2023-10-05T09:35:01.6612492Z _get_conv_bn_getitem_nodes(r.replacements) 2023-10-05T09:35:01.6613058Z File "/opt/conda/envs/py_3.8/lib/python3.8/site-packages/torch/ao/quantization/pt2e/qat_utils.py", line 383, in _get_conv_bn_getitem_nodes 2023-10-05T09:35:01.6613465Z assert bn_node is not None 2023-10-05T09:35:01.6613716Z AssertionError ``` Pull Request resolved: #110653 Approved by: https://github.com/jerryzh168, https://github.com/pruthvistony

andrewor14 requested a review from jerryzh168 as a code owner September 22, 2023 20:12

pytorch-bot bot added the release notes: quantization release notes category label Sep 22, 2023

andrewor14 requested a review from kimishpatel September 22, 2023 20:12

andrewor14 mentioned this pull request Sep 22, 2023

[quant][pt2] Fix conv bn decomposition in cuda mode #108809

Closed

jerryzh168 approved these changes Sep 22, 2023

View reviewed changes

pytorch-bot bot added the ciflow/trunk Trigger trunk jobs on your pull request label Oct 5, 2023

pytorchmergebot added the merging label Oct 5, 2023

pytorchmergebot added Merged and removed merging labels Oct 5, 2023

pytorchmergebot closed this in 62cad5b Oct 5, 2023

jeffdaily mentioned this pull request Oct 5, 2023

[quant][pt2][ROCm] follow-up PR 109908 for miopen_batch_norm #110653

Closed

facebook-github-bot deleted the gh/andrewor14/32/head branch October 8, 2023 14:22

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

[quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

Uh oh!

andrewor14 commented Sep 22, 2023 •

edited

Loading

Uh oh!

pytorch-bot bot commented Sep 22, 2023 •

edited

Loading

Uh oh!

jerryzh168 left a comment

Uh oh!

andrewor14 commented Sep 25, 2023

Uh oh!

andrewor14 commented Sep 27, 2023

Uh oh!

andrewor14 commented Sep 28, 2023

Uh oh!

andrewor14 commented Oct 4, 2023

Uh oh!

andrewor14 commented Oct 5, 2023

Uh oh!

pytorchmergebot commented Oct 5, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

[quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

[quant][pt2] Support cudnn_batch_norm in QAT fusion #109908

Uh oh!

Conversation

andrewor14 commented Sep 22, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented Sep 22, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/109908

✅ You can merge normally! (3 Unrelated Failures)

Uh oh!

jerryzh168 left a comment

Choose a reason for hiding this comment

Uh oh!

andrewor14 commented Sep 25, 2023

Uh oh!

andrewor14 commented Sep 27, 2023

Uh oh!

andrewor14 commented Sep 28, 2023

Uh oh!

andrewor14 commented Oct 4, 2023

Uh oh!

andrewor14 commented Oct 5, 2023

Uh oh!

pytorchmergebot commented Oct 5, 2023

Merge started

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

andrewor14 commented Sep 22, 2023 •

edited

Loading

pytorch-bot bot commented Sep 22, 2023 •

edited

Loading