BFloat16: enable prepacked weights's inference #37218

XiaobingSuper · 2020-04-24T09:08:29Z

Stack from ghstack:

BFloat16: enable prepacked weights's inference #37218 BFloat16: enable prepacked weights's inference
BFloat16: add explicit dtype support for to_mkldnn and to_dense #37215 BFloat16: add explicit dtype support for to_mkldnn and to_dense
add mkldnn relu backward and reshape backward #37199 add mkldnn relu backward and reshape backward
add mkldnn batch_norm backward #37147 add mkldnn batch_norm backward
add mkldnn pooling backward #37146 add mkldnn pooling backward
add mkldnn linear backward #36122 add mkldnn linear backward
enable mkldnn conv2d backward to support mkldnn tensor input #36121 enable mkldnn conv2d backward to support mkldnn tensor input

Differential Revision: D22440970

[ghstack-poisoned]

ghstack-source-id: 78316d4 Pull Request resolved: #37218

XiaobingSuper · 2020-04-24T09:17:41Z

@jgong5, @hongzhen1

dr-ci · 2020-04-24T11:04:50Z

💊 CI failures summary and remediations

As of commit cd4fda6 (more details on the Dr. CI page):

6/7 failures introduced in this PR
1/7 tentatively recognized as flaky ❄️
- Click here to rerun these jobs

🕵️ 6 new failures recognized by patterns

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

pytorch_windows_vs2019_py36_cuda10.1_test2 (1/6)

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

RuntimeError: test_mkldnn failed!

 
FAILED (errors=2, skipped=3) 
 
Generating XML reports... 
Generated XML report: test-reports\python-unittest\TEST-TestMkldnn-20200708232437.xml 
Traceback (most recent call last): 
  File "run_test.py", line 728, in <module> 
    main() 
  File "run_test.py", line 721, in main 
    raise RuntimeError(message) 
RuntimeError: test_mkldnn failed! 
 
(base) circleci@PACKER-5EFB90C2 C:\Users\circleci\project\test>if ERRORLEVEL 1 exit /b 1  
+ cleanup
+ retcode=1
+ set +x

pytorch_linux_bionic_py3_6_clang9_test (2/6)

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

Jul 08 23:55:17 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" }

Jul 08 23:55:17     raise RuntimeError(message) 
Jul 08 23:55:17 RuntimeError: test_mkldnn failed! 
Jul 08 23:55:17  
Jul 08 23:55:17 real	16m11.849s 
Jul 08 23:55:17 user	12m20.316s 
Jul 08 23:55:17 sys	2m21.501s 
Jul 08 23:55:17 + cleanup 
Jul 08 23:55:17 + retcode=1 
Jul 08 23:55:17 + set +x 
Jul 08 23:55:17 =================== sccache compilation log =================== 
Jul 08 23:55:17 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" } 
Jul 08 23:55:17  
Jul 08 23:55:17 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 08 23:55:17 Compile requests                 65 
Jul 08 23:55:17 Compile requests executed        35 
Jul 08 23:55:17 Cache hits                       15 
Jul 08 23:55:17 Cache misses                     19 
Jul 08 23:55:17 Cache timeouts                    0 
Jul 08 23:55:17 Cache read errors                 0 
Jul 08 23:55:17 Forced recaches                   0 
Jul 08 23:55:17 Cache write errors                0

pytorch_linux_xenial_py3_6_gcc5_4_ge_config_simple_test (3/6)

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

Jul 08 23:56:23 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" }

Jul 08 23:56:23 Traceback (most recent call last): 
Jul 08 23:56:23   File "test/run_test.py", line 728, in <module> 
Jul 08 23:56:23     main() 
Jul 08 23:56:23   File "test/run_test.py", line 721, in main 
Jul 08 23:56:23     raise RuntimeError(message) 
Jul 08 23:56:23 RuntimeError: test_mkldnn failed! 
Jul 08 23:56:23 + cleanup 
Jul 08 23:56:23 + retcode=1 
Jul 08 23:56:23 + set +x 
Jul 08 23:56:23 =================== sccache compilation log =================== 
Jul 08 23:56:23 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" } 
Jul 08 23:56:23  
Jul 08 23:56:23 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 08 23:56:23 Compile requests                 65 
Jul 08 23:56:23 Compile requests executed        35 
Jul 08 23:56:23 Cache hits                        3 
Jul 08 23:56:23 Cache misses                     31 
Jul 08 23:56:23 Cache timeouts                    0 
Jul 08 23:56:23 Cache read errors                 0 
Jul 08 23:56:23 Forced recaches                   0 
Jul 08 23:56:23 Cache write errors                0

pytorch_linux_xenial_py3_6_gcc5_4_test (4/6)

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

Jul 08 23:56:43 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" }

Jul 08 23:56:42 Traceback (most recent call last): 
Jul 08 23:56:42   File "test/run_test.py", line 728, in <module> 
Jul 08 23:56:42     main() 
Jul 08 23:56:42   File "test/run_test.py", line 721, in main 
Jul 08 23:56:42     raise RuntimeError(message) 
Jul 08 23:56:42 RuntimeError: test_mkldnn failed! 
Jul 08 23:56:43 + cleanup 
Jul 08 23:56:43 + retcode=1 
Jul 08 23:56:43 + set +x 
Jul 08 23:56:43 =================== sccache compilation log =================== 
Jul 08 23:56:43 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" } 
Jul 08 23:56:43  
Jul 08 23:56:43 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 08 23:56:43 Compile requests                 65 
Jul 08 23:56:43 Compile requests executed        35 
Jul 08 23:56:43 Cache hits                       15 
Jul 08 23:56:43 Cache misses                     19 
Jul 08 23:56:43 Cache timeouts                    0 
Jul 08 23:56:43 Cache read errors                 0 
Jul 08 23:56:43 Forced recaches                   0 
Jul 08 23:56:43 Cache write errors                0

pytorch_linux_bionic_py3_8_gcc9_test (5/6)

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

Jul 09 00:23:36 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function ‘int main()’:\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:22: error: expected ‘;’ before ‘}’ token\n    2 | int main() { return 0 }\n      |                      ^~\n      |                      ;\n" }

Jul 09 00:23:36     raise RuntimeError(message) 
Jul 09 00:23:36 RuntimeError: test_mkldnn failed! 
Jul 09 00:23:36  
Jul 09 00:23:36 real	19m24.999s 
Jul 09 00:23:36 user	11m22.310s 
Jul 09 00:23:36 sys	0m48.862s 
Jul 09 00:23:36 + cleanup 
Jul 09 00:23:36 + retcode=1 
Jul 09 00:23:36 + set +x 
Jul 09 00:23:36 =================== sccache compilation log =================== 
Jul 09 00:23:36 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function ‘int main()’:\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:22: error: expected ‘;’ before ‘}’ token\n    2 | int main() { return 0 }\n      |                      ^~\n      |                      ;\n" } 
Jul 09 00:23:36  
Jul 09 00:23:36 =========== If your build fails, please take a look at the log above for possible reasons =========== 
Jul 09 00:23:36 Compile requests                 65 
Jul 09 00:23:36 Compile requests executed        35 
Jul 09 00:23:36 Cache hits                        2 
Jul 09 00:23:36 Cache misses                     32 
Jul 09 00:23:36 Cache timeouts                    0 
Jul 09 00:23:36 Cache read errors                 0 
Jul 09 00:23:36 Forced recaches                   0 
Jul 09 00:23:36 Cache write errors                0

pytorch_macos_10_13_py3_test (6/6)

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

Jul 08 17:19:05 RuntimeError: test_mkldnn failed!

Jul 08 17:19:05  
Jul 08 17:19:05 FAILED (errors=2, skipped=3) 
Jul 08 17:19:05  
Jul 08 17:19:05 Generating XML reports... 
Jul 08 17:19:05 Generated XML report: test-reports/dist-gloo/TEST-TestMkldnn-20200708171902.xml 
Jul 08 17:19:05 Traceback (most recent call last): 
Jul 08 17:19:05   File "test/run_test.py", line 728, in <module> 
Jul 08 17:19:05     main() 
Jul 08 17:19:05   File "test/run_test.py", line 721, in main 
Jul 08 17:19:05     raise RuntimeError(message) 
Jul 08 17:19:05 RuntimeError: test_mkldnn failed! 
Jul 08 17:19:05 + cleanup 
Jul 08 17:19:05 + retcode=1 
Jul 08 17:19:05 + set +x

❄️ 1 failure tentatively classified as flaky

but reruns have not yet been triggered to confirm:

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test (1/1)

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

Jul 09 02:15:05 ConnectionResetError: [Errno 104] Connection reset by peer

Jul 09 02:15:05   File "/opt/conda/lib/python3.6/multiprocessing/connection.py", line 455, in accept 
Jul 09 02:15:05     deliver_challenge(c, self._authkey) 
Jul 09 02:15:05   File "/opt/conda/lib/python3.6/multiprocessing/connection.py", line 722, in deliver_challenge 
Jul 09 02:15:05     response = connection.recv_bytes(256)        # reject large message 
Jul 09 02:15:05   File "/opt/conda/lib/python3.6/multiprocessing/connection.py", line 216, in recv_bytes 
Jul 09 02:15:05     buf = self._recv_bytes(maxlength) 
Jul 09 02:15:05   File "/opt/conda/lib/python3.6/multiprocessing/connection.py", line 407, in _recv_bytes 
Jul 09 02:15:05     buf = self._recv(4) 
Jul 09 02:15:05   File "/opt/conda/lib/python3.6/multiprocessing/connection.py", line 379, in _recv 
Jul 09 02:15:05     chunk = read(handle, remaining) 
Jul 09 02:15:05 ConnectionResetError: [Errno 104] Connection reset by peer 
Jul 09 02:15:05 /opt/conda/lib/python3.6/multiprocessing/semaphore_tracker.py:143: UserWarning: semaphore_tracker: There appear to be 14 leaked semaphores to clean up at shutdown 
Jul 09 02:15:05   len(cache)) 
Jul 09 02:15:07 Process ErrorTrackingProcess-122: 
Jul 09 02:15:07 Traceback (most recent call last): 
Jul 09 02:15:07   File "/opt/conda/lib/python3.6/multiprocessing/process.py", line 258, in _bootstrap 
Jul 09 02:15:07     self.run() 
Jul 09 02:15:07   File "/var/lib/jenkins/workspace/test/test_dataloader.py", line 361, in run 
Jul 09 02:15:07     super(ErrorTrackingProcess, self).run() 
Jul 09 02:15:07   File "/opt/conda/lib/python3.6/multiprocessing/process.py", line 93, in run 
Jul 09 02:15:07     self._target(*self._args, **self._kwargs)

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 98 times.

[ghstack-poisoned]

ghstack-source-id: c8590a5 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 03d5d1f Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 107e0aa Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 77a2856 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 162c824 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 35d85db Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 81594b6 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 7557846 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 2cfb739 Pull Request resolved: #37218

[ghstack-poisoned]

ghstack-source-id: 02994af Pull Request resolved: #37218

Differential Revision: [D22440970](https://our.internmc.facebook.com/intern/diff/D22440970) [ghstack-poisoned]

ghstack-source-id: 76bb2a8 Pull Request resolved: #37218

facebook-github-bot · 2020-10-30T17:25:56Z

Hi @XiaobingSuper!

Thank you for your pull request. We require contributors to sign our Contributor License Agreement, and yours needs attention.

You currently have a record in our system, but we do not have a signature on file.

In order for us to review and merge your code, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

If you have received this in error or have any questions, please contact us at [email protected]. Thanks!

VitalyFedyunin · 2020-11-17T21:51:52Z

Hello @XiaobingSuper can be please tag necessary ideep version for this PR. See more in #47853

BFloat16: enable prepacked weights's inference

1184f28

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Apr 24, 2020

BFloat16: enable prepacked weights's inference

c7e8b7b

ghstack-source-id: 78316d4 Pull Request resolved: #37218

XiaobingSuper requested review from VitalyFedyunin and pinzhenx April 24, 2020 09:13

pytorchbot added the open source label Apr 24, 2020

XiaobingSuper requested a review from albanD April 26, 2020 02:19

Update on "BFloat16: enable prepacked weights's inference"

f11f084

[ghstack-poisoned]

Update on "BFloat16: enable prepacked weights's inference"

f626718

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Apr 26, 2020

BFloat16: enable prepacked weights's inference

320e821

ghstack-source-id: c8590a5 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

df594f5

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Apr 26, 2020

BFloat16: enable prepacked weights's inference

00175a2

ghstack-source-id: 03d5d1f Pull Request resolved: #37218

XiaobingSuper added module: mkldnn Related to Intel IDEEP or oneDNN (a.k.a. mkldnn) integration module: bfloat16 labels Apr 27, 2020

Update on "BFloat16: enable prepacked weights's inference"

4965b33

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Apr 29, 2020

BFloat16: enable prepacked weights's inference

5180ea7

ghstack-source-id: 107e0aa Pull Request resolved: #37218

ngimel added the triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module label May 5, 2020

Update on "BFloat16: enable prepacked weights's inference"

e7da1e3

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request May 7, 2020

BFloat16: enable prepacked weights's inference

1d4f3b4

ghstack-source-id: 77a2856 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

978b70d

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request May 9, 2020

BFloat16: enable prepacked weights's inference

7727a6f

ghstack-source-id: 162c824 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

b9a3dad

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request May 9, 2020

BFloat16: enable prepacked weights's inference

d1a247e

ghstack-source-id: 35d85db Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

9151ba0

[ghstack-poisoned]

Update on "BFloat16: enable prepacked weights's inference"

ed152af

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Jun 22, 2020

BFloat16: enable prepacked weights's inference

c2f612c

ghstack-source-id: 81594b6 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

abb91f6

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Jun 28, 2020

BFloat16: enable prepacked weights's inference

a260f5b

ghstack-source-id: 7557846 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

04562a4

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Jul 2, 2020

BFloat16: enable prepacked weights's inference

551725a

ghstack-source-id: 2cfb739 Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

63df808

[ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Jul 3, 2020

BFloat16: enable prepacked weights's inference

1d89cb9

ghstack-source-id: 02994af Pull Request resolved: #37218

Update on "BFloat16: enable prepacked weights's inference"

4d171ed

Differential Revision: [D22440970](https://our.internmc.facebook.com/intern/diff/D22440970) [ghstack-poisoned]

Update on "BFloat16: enable prepacked weights's inference"

cd4fda6

Differential Revision: [D22440970](https://our.internmc.facebook.com/intern/diff/D22440970) [ghstack-poisoned]

XiaobingSuper added a commit that referenced this pull request Jul 8, 2020

BFloat16: enable prepacked weights's inference

458ce5d

ghstack-source-id: 76bb2a8 Pull Request resolved: #37218

facebook-github-bot added the cla signed label Nov 11, 2020

VitalyFedyunin approved these changes Dec 14, 2020

View reviewed changes

XiaobingSuper closed this Dec 16, 2020

facebook-github-bot deleted the gh/xiaobingsuper/18/head branch January 15, 2021 15:17

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

BFloat16: enable prepacked weights's inference #37218

BFloat16: enable prepacked weights's inference #37218

Uh oh!

XiaobingSuper commented Apr 24, 2020 •

edited by VitalyFedyunin

Loading

Uh oh!

XiaobingSuper commented Apr 24, 2020

Uh oh!

dr-ci bot commented Apr 24, 2020 •

edited

Loading

Uh oh!

facebook-github-bot commented Oct 30, 2020

Uh oh!

VitalyFedyunin commented Nov 17, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

BFloat16: enable prepacked weights's inference #37218

BFloat16: enable prepacked weights's inference #37218

Uh oh!

Conversation

XiaobingSuper commented Apr 24, 2020 • edited by VitalyFedyunin Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

XiaobingSuper commented Apr 24, 2020

Uh oh!

dr-ci bot commented Apr 24, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🕵️ 6 new failures recognized by patterns

pytorch_windows_vs2019_py36_cuda10.1_test2 (1/6)

pytorch_linux_bionic_py3_6_clang9_test (2/6)

pytorch_linux_xenial_py3_6_gcc5_4_ge_config_simple_test (3/6)

pytorch_linux_xenial_py3_6_gcc5_4_test (4/6)

pytorch_linux_bionic_py3_8_gcc9_test (5/6)

pytorch_macos_10_13_py3_test (6/6)

❄️ 1 failure tentatively classified as flaky

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test (1/1)

Uh oh!

facebook-github-bot commented Oct 30, 2020

Uh oh!

VitalyFedyunin commented Nov 17, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

XiaobingSuper commented Apr 24, 2020 •

edited by VitalyFedyunin

Loading

dr-ci bot commented Apr 24, 2020 •

edited

Loading