Skip to content

[quant][graphmode] Refactor quantization patterns #40894

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
supriyar wants to merge 4 commits into from
Closed

[quant][graphmode] Refactor quantization patterns #40894

supriyar wants to merge 4 commits into from

Conversation

@supriyar
Copy link
Contributor

@supriyar supriyar commented Jul 2, 2020 •
edited
Loading

Stack from ghstack:

Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

Differential Revision: D22403901

@supriyar
[quant][graphmode] Refactor quantization patterns
97b1fcc 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

[ghstack-poisoned]
@supriyar supriyar requested a review from apaszke as a code owner July 2, 2020 02:06
supriyar added a commit that referenced this pull request Jul 2, 2020
@supriyar
[quant][graphmode] Refactor quantization patterns
af3b835 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

ghstack-source-id: 88aa426
Pull Request resolved: #40894
@facebook-github-bot facebook-github-bot added the oncall: jit Add this issue/PR to JIT oncall triage queue label Jul 2, 2020
@dr-ci
Copy link

dr-ci bot commented Jul 2, 2020 •
edited
Loading

💊 CI failures summary and remediations

As of commit 73794d4 (more details on the Dr. CI page):

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

See CircleCI build pytorch_macos_10_13_py3_test (1/1)

Step: "Test" (full log | diagnosis details | 🔁 rerun)

Jul 07 12:52:54 [E request_callback_impl.cpp:168] Received error while processing request type 2: PickleError: ScriptModules cannot be deepcopied using copy.deepcopy or saved using torch.save. Mixed serialization of script and non-script modules is not supported. For purely script modules use my_script_module.save() instead. 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(93): serialize 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(145): serialize 
Jul 07 12:52:54  
Jul 07 12:52:54 [E request_callback_impl.cpp:168] Received error while processing request type 2: PickleError: ScriptModules cannot be deepcopied using copy.deepcopy or saved using torch.save. Mixed serialization of script and non-script modules is not supported. For purely script modules use my_script_module.save(<filename>) instead. 
Jul 07 12:52:54  
Jul 07 12:52:54 At: 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/jit/__init__.py(1149): __getstate__ 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(93): serialize 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(145): serialize 
Jul 07 12:52:54  
Jul 07 12:52:54 [E request_callback_impl.cpp:168] Received error while processing request type 2: PickleError: ScriptModules cannot be deepcopied using copy.deepcopy or saved using torch.save. Mixed serialization of script and non-script modules is not supported. For purely script modules use my_script_module.save(<filename>) instead. 
Jul 07 12:52:54  
Jul 07 12:52:54 At: 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/jit/__init__.py(1149): __getstate__ 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(93): serialize 
Jul 07 12:52:54   /Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/distributed/rpc/internal.py(145): serialize 
Jul 07 12:52:54  
Jul 07 12:52:55 ok (1.319s) 
Jul 07 12:52:56   test_unexepected_kwarg_is_specified (__main__.JitRpcTestWithSpawn) ... ok (1.200s) 
Jul 07 12:52:57   test_user_rrefs_confirmed (__main__.JitRpcTestWithSpawn) ... ok (1.189s) 
Jul 07 12:52:58   test_user_rrefs_confirmed_remote (__main__.JitRpcTestWithSpawn) ... ok (1.270s)

❄️ 2 failures tentatively classified as flaky

but reruns have not yet been triggered to confirm:

See CircleCI build pytorch_linux_xenial_cuda10_2_cudnn7_py3_ge_config_profiling_test (1/2)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun) ❄️

Jul 07 18:06:16 E: Failed to fetch https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/Packages Writing more data than expected (1196030 > 1190062) 
Jul 07 18:06:11 + echo 'Testing pytorch' 
Jul 07 18:06:11 + '[' -n 1 ']' 
Jul 07 18:06:11 + pip_install unittest-xml-reporting 
Jul 07 18:06:11 + pip install --progress-bar off unittest-xml-reporting 
Jul 07 18:06:12 Collecting unittest-xml-reporting 
Jul 07 18:06:12   Downloading unittest_xml_reporting-3.0.2-py2.py3-none-any.whl (19 kB) 
Jul 07 18:06:12 Installing collected packages: unittest-xml-reporting 
Jul 07 18:06:12 Successfully installed unittest-xml-reporting-3.0.2 
Jul 07 18:06:12 + [[ pytorch-linux-xenial-cuda10.1-cudnn7-ge_config_profiling-test == *-xenial-cuda10.1-* ]] 
Jul 07 18:06:12 + sudo apt-get -qq update 
Jul 07 18:06:16 E: Failed to fetch https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/Packages  Writing more data than expected (1196030 > 1190062) 
Jul 07 18:06:16 E: Some index files failed to download. They have been ignored, or old ones used instead. 
Jul 07 18:06:16 + cleanup 
Jul 07 18:06:16 + retcode=100 
Jul 07 18:06:16 + set +x 
Jul 07 18:06:16 =================== sccache compilation log =================== 
Jul 07 18:06:16 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 07 18:06:16 Compile requests                 0 
Jul 07 18:06:16 Compile requests executed        0 
Jul 07 18:06:16 Cache hits                       0 
Jul 07 18:06:16 Cache misses                     0

See CircleCI build pytorch_linux_xenial_cuda10_2_cudnn7_py3_ge_config_legacy_test (2/2)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun) ❄️

Jul 07 18:05:20 E: Failed to fetch https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/Packages Writing more data than expected (1196031 > 1190062) 
Jul 07 18:05:15 + echo 'Testing pytorch' 
Jul 07 18:05:15 + '[' -n 1 ']' 
Jul 07 18:05:15 + pip_install unittest-xml-reporting 
Jul 07 18:05:15 + pip install --progress-bar off unittest-xml-reporting 
Jul 07 18:05:16 Collecting unittest-xml-reporting 
Jul 07 18:05:16   Downloading unittest_xml_reporting-3.0.2-py2.py3-none-any.whl (19 kB) 
Jul 07 18:05:16 Installing collected packages: unittest-xml-reporting 
Jul 07 18:05:16 Successfully installed unittest-xml-reporting-3.0.2 
Jul 07 18:05:16 + [[ pytorch-linux-xenial-cuda10.1-cudnn7-ge_config_legacy-test == *-xenial-cuda10.1-* ]] 
Jul 07 18:05:16 + sudo apt-get -qq update 
Jul 07 18:05:20 E: Failed to fetch https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/Packages  Writing more data than expected (1196031 > 1190062) 
Jul 07 18:05:20 E: Some index files failed to download. They have been ignored, or old ones used instead. 
Jul 07 18:05:20 + cleanup 
Jul 07 18:05:20 + retcode=100 
Jul 07 18:05:20 + set +x 
Jul 07 18:05:20 =================== sccache compilation log =================== 
Jul 07 18:05:20 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 07 18:05:20 Compile requests                 0 
Jul 07 18:05:20 Compile requests executed        0 
Jul 07 18:05:20 Cache hits                       0 
Jul 07 18:05:20 Cache misses                     0
This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group.

See how this bot performed.

This comment has been revised 9 times.

@supriyar
Update on "[quant][graphmode] Refactor quantization patterns"
611f11d 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

[ghstack-poisoned]
@supriyar supriyar requested review from jerryzh168 and vkuzo July 2, 2020 17:54
supriyar added a commit that referenced this pull request Jul 2, 2020
@supriyar
[quant][graphmode] Refactor quantization patterns
06a71ce 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

ghstack-source-id: 19f1f63
Pull Request resolved: #40894
jerryzh168
jerryzh168 reviewed Jul 6, 2020
View reviewed changes
torch/csrc/jit/passes/quantization/quantization_patterns.h
return (%r) )";

return {
{"conv1d_prepack_unpack", conv1d_with_quant, conv1d_with_quant_prepack},
Copy link
Contributor

@jerryzh168 jerryzh168 Jul 6, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks like quantized op is not used actually, I'm also thinking of removing this field and we can make the struct more general to all users of subgraph rewriter

jerryzh168
jerryzh168 approved these changes Jul 6, 2020
View reviewed changes
Copy link
Contributor

@jerryzh168 jerryzh168 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks

@supriyar
Update on "[quant][graphmode] Refactor quantization patterns"
61fef6a 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

[ghstack-poisoned]
@supriyar supriyar mentioned this pull request Jul 7, 2020
Closed
@supriyar
Update on "[quant][graphmode] Refactor quantization patterns"
73794d4 
Summary:

Test Plan:
python test/test_quantization.py

Reviewers:

Subscribers:

Tasks:

Tags:

Differential Revision: [D22403901](https://our.internmc.facebook.com/intern/diff/D22403901)

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

facebook-github-bot commented Jul 8, 2020

This pull request has been merged in dde1804.

@facebook-github-bot facebook-github-bot deleted the gh/supriyar/147/head branch July 12, 2020 14:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@vkuzo vkuzo vkuzo approved these changes

@jerryzh168 jerryzh168 jerryzh168 approved these changes

@apaszke apaszke Awaiting requested review from apaszke

Assignees

No one assigned

Labels

Merged oncall: jit Add this issue/PR to JIT oncall triage queue

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

6 participants

@supriyar @facebook-github-bot @vkuzo @jerryzh168 @mruberry