[Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future by wayi1 · Pull Request #51094 · pytorch/pytorch

wayi1 · 2021-01-26T06:34:22Z

Stack from ghstack:

[Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future #51094 [Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future
[Gradient Compression] Explicitly specify the dtype of the error tensor #50985 [Gradient Compression] Explicitly specify the dtype of the error tensor
[Gradient Compression] Simplify the implementation of warm-start #50981 [Gradient Compression] Simplify the implementation of warm-start
[Gradient Compression] Typo fixes in PowerSGD #50974 [Gradient Compression] Typo fixes in PowerSGD
[Gradient Compression] Allow PowerSGD to run vallina allreduce for the first K iterations #50973 [Gradient Compression] Allow PowerSGD to run vallina allreduce for the first K iterations

Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202

Differential Revision: D26070147

… by creating a util function that make a vanilla allreduce future Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26070147](https://our.internmc.facebook.com/intern/diff/D26070147/) [ghstack-poisoned]

facebook-github-bot · 2021-01-26T06:34:28Z

💊 CI failures summary and remediations

As of commit 2ca21e5 (more details on the Dr. CI page):

9/9 failures introduced in this PR

🕵️ 9 new failures recognized by patterns

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

pytorch_linux_xenial_py3_clang5_asan_test2 (1/9)

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

Jan 29 00:53:48 SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /var/lib/jenkins/workspace/aten/src/ATen/Utils.cpp:15:3 in

Jan 29 00:53:48     #7 0x560ac09b870b in PyEval_EvalCode /tmp/build/80754af9/python_1599604603603/work/Python/ceval.c:731
Jan 29 00:53:48     #8 0x560ac0a38573 in run_mod /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:1025
Jan 29 00:53:48     #9 0x560ac0a3860c in PyRun_StringFlags /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:949
Jan 29 00:53:48     #10 0x560ac0a3866e in PyRun_SimpleStringFlags /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:445
Jan 29 00:53:48     #11 0x560ac0a3c472 in run_command /tmp/build/80754af9/python_1599604603603/work/Modules/main.c:301
Jan 29 00:53:48     #12 0x560ac0a3c472 in Py_Main /tmp/build/80754af9/python_1599604603603/work/Modules/main.c:749
Jan 29 00:53:48     #13 0x560ac090643d in main /tmp/build/80754af9/python_1599604603603/work/Programs/python.c:69
Jan 29 00:53:48     #14 0x7ffa1d8e283f in __libc_start_main /build/glibc-e6zv40/glibc-2.23/csu/../csu/libc-start.c:291
Jan 29 00:53:48     #15 0x560ac09e5d0a in _start /home/rdonnelly/mc/conda-bld/compilers_linux-64_1534865402226/work/.build/src/glibc-2.12.2/csu/../sysdeps/x86_64/elf/start.S:103
Jan 29 00:53:48 
Jan 29 00:53:48 SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /var/lib/jenkins/workspace/aten/src/ATen/Utils.cpp:15:3 in 
Jan 29 00:53:48 + retcode=1
Jan 29 00:53:48 + set -e
Jan 29 00:53:48 + return 1
Jan 29 00:53:48 + [[ pytorch-linux-xenial-py3-clang5-asan-test2 == *-NO_AVX-* ]]
Jan 29 00:53:48 + [[ pytorch-linux-xenial-py3-clang5-asan-test2 == *-NO_AVX2-* ]]
Jan 29 00:53:48 + '[' -n https://github.com/pytorch/pytorch/pull/51094 ']'
Jan 29 00:53:48 + [[ pytorch-linux-xenial-py3-clang5-asan-test2 != *coverage* ]]
Jan 29 00:53:48 ++ mktemp
Jan 29 00:53:48 + DETERMINE_FROM=/tmp/tmp.fpKEWaHwLo
Jan 29 00:53:48 + file_diff_from_base /tmp/tmp.fpKEWaHwLo

pytorch_linux_bionic_py3_6_clang9_test (2/9)

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

Jan 29 00:52:06 sccache: error: couldn't connect to server

Jan 29 00:52:06 +++ eval 'extract_trap_cmd '
Jan 29 00:52:06 ++++ extract_trap_cmd
Jan 29 00:52:06 ++++ printf '%s\n' ''
Jan 29 00:52:06 +++ printf '%s\n' cleanup
Jan 29 00:52:06 ++ trap -- '
Jan 29 00:52:06 cleanup' EXIT
Jan 29 00:52:06 ++ [[ pytorch-linux-bionic-py3.6-clang9-test != *pytorch-win-* ]]
Jan 29 00:52:06 ++ which sccache
Jan 29 00:52:06 ++ sccache --stop-server
Jan 29 00:52:06 Stopping sccache server...
Jan 29 00:52:06 sccache: error: couldn't connect to server
Jan 29 00:52:06 sccache: caused by: Connection refused (os error 111)
Jan 29 00:52:06 ++ true
Jan 29 00:52:06 ++ rm /var/lib/jenkins/sccache_error.log
Jan 29 00:52:06 ++ [[ pytorch-linux-bionic-py3.6-clang9-test == *rocm* ]]
Jan 29 00:52:06 ++ SCCACHE_ERROR_LOG=/var/lib/jenkins/sccache_error.log
Jan 29 00:52:06 ++ SCCACHE_IDLE_TIMEOUT=1200
Jan 29 00:52:06 ++ RUST_LOG=sccache::server=error
Jan 29 00:52:06 ++ sccache --start-server
Jan 29 00:52:06 sccache: Starting the server...
Jan 29 00:52:06 ++ sccache --zero-stats

pytorch_xla_linux_bionic_py3_6_clang9_build (3/9)

Step: "(Optional) Merge target branch" (full log | diagnosis details | 🔁 rerun)

Automatic merge failed; fix conflicts and then commit the result.

CONFLICT (add/add): Merge conflict in .jenkins/caffe2/test.sh
Auto-merging .jenkins/caffe2/test.sh
CONFLICT (add/add): Merge conflict in .gitmodules
Auto-merging .gitmodules
CONFLICT (add/add): Merge conflict in .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
Auto-merging .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
CONFLICT (add/add): Merge conflict in .circleci/scripts/python_doc_push_script.sh
Auto-merging .circleci/scripts/python_doc_push_script.sh
CONFLICT (add/add): Merge conflict in .circleci/config.yml
Auto-merging .circleci/config.yml
Automatic merge failed; fix conflicts and then commit the result.


Exited with code exit status 1

pytorch_linux_xenial_py3_6_gcc5_4_build (4/9)

Step: "(Optional) Merge target branch" (full log | diagnosis details | 🔁 rerun)

Automatic merge failed; fix conflicts and then commit the result.

CONFLICT (add/add): Merge conflict in .jenkins/caffe2/test.sh
Auto-merging .jenkins/caffe2/test.sh
CONFLICT (add/add): Merge conflict in .gitmodules
Auto-merging .gitmodules
CONFLICT (add/add): Merge conflict in .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
Auto-merging .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
CONFLICT (add/add): Merge conflict in .circleci/scripts/python_doc_push_script.sh
Auto-merging .circleci/scripts/python_doc_push_script.sh
CONFLICT (add/add): Merge conflict in .circleci/config.yml
Auto-merging .circleci/config.yml
Automatic merge failed; fix conflicts and then commit the result.


Exited with code exit status 1

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test2 (5/9)

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

Jan 29 01:12:31 sccache: error: couldn't connect to server

Jan 29 01:12:31 +++ eval 'extract_trap_cmd '
Jan 29 01:12:31 ++++ extract_trap_cmd
Jan 29 01:12:31 ++++ printf '%s\n' ''
Jan 29 01:12:31 +++ printf '%s\n' cleanup
Jan 29 01:12:31 ++ trap -- '
Jan 29 01:12:31 cleanup' EXIT
Jan 29 01:12:31 ++ [[ pytorch-linux-xenial-cuda10.2-cudnn7-py3-gcc7-test2 != *pytorch-win-* ]]
Jan 29 01:12:31 ++ which sccache
Jan 29 01:12:31 ++ sccache --stop-server
Jan 29 01:12:31 Stopping sccache server...
Jan 29 01:12:31 sccache: error: couldn't connect to server
Jan 29 01:12:31 sccache: caused by: Connection refused (os error 111)
Jan 29 01:12:31 ++ true
Jan 29 01:12:31 ++ rm /var/lib/jenkins/sccache_error.log
Jan 29 01:12:31 ++ [[ pytorch-linux-xenial-cuda10.2-cudnn7-py3-gcc7-test2 == *rocm* ]]
Jan 29 01:12:31 ++ SCCACHE_ERROR_LOG=/var/lib/jenkins/sccache_error.log
Jan 29 01:12:31 ++ SCCACHE_IDLE_TIMEOUT=1200
Jan 29 01:12:31 ++ RUST_LOG=sccache::server=error
Jan 29 01:12:31 ++ sccache --start-server
Jan 29 01:12:31 sccache: Starting the server...
Jan 29 01:12:31 ++ sccache --zero-stats

pytorch_linux_xenial_py3_clang5_asan_test1 (6/9)

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

Jan 29 00:53:45 SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /var/lib/jenkins/workspace/aten/src/ATen/Utils.cpp:15:3 in

Jan 29 00:53:45     #7 0x56041871170b in PyEval_EvalCode /tmp/build/80754af9/python_1599604603603/work/Python/ceval.c:731
Jan 29 00:53:45     #8 0x560418791573 in run_mod /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:1025
Jan 29 00:53:45     #9 0x56041879160c in PyRun_StringFlags /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:949
Jan 29 00:53:45     #10 0x56041879166e in PyRun_SimpleStringFlags /tmp/build/80754af9/python_1599604603603/work/Python/pythonrun.c:445
Jan 29 00:53:45     #11 0x560418795472 in run_command /tmp/build/80754af9/python_1599604603603/work/Modules/main.c:301
Jan 29 00:53:45     #12 0x560418795472 in Py_Main /tmp/build/80754af9/python_1599604603603/work/Modules/main.c:749
Jan 29 00:53:45     #13 0x56041865f43d in main /tmp/build/80754af9/python_1599604603603/work/Programs/python.c:69
Jan 29 00:53:45     #14 0x7fa9a753183f in __libc_start_main /build/glibc-e6zv40/glibc-2.23/csu/../csu/libc-start.c:291
Jan 29 00:53:45     #15 0x56041873ed0a in _start /home/rdonnelly/mc/conda-bld/compilers_linux-64_1534865402226/work/.build/src/glibc-2.12.2/csu/../sysdeps/x86_64/elf/start.S:103
Jan 29 00:53:45 
Jan 29 00:53:45 SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /var/lib/jenkins/workspace/aten/src/ATen/Utils.cpp:15:3 in 
Jan 29 00:53:45 + retcode=1
Jan 29 00:53:45 + set -e
Jan 29 00:53:45 + return 1
Jan 29 00:53:45 + [[ pytorch-linux-xenial-py3-clang5-asan-test1 == *-NO_AVX-* ]]
Jan 29 00:53:45 + [[ pytorch-linux-xenial-py3-clang5-asan-test1 == *-NO_AVX2-* ]]
Jan 29 00:53:45 + '[' -n https://github.com/pytorch/pytorch/pull/51094 ']'
Jan 29 00:53:45 + [[ pytorch-linux-xenial-py3-clang5-asan-test1 != *coverage* ]]
Jan 29 00:53:45 ++ mktemp
Jan 29 00:53:45 + DETERMINE_FROM=/tmp/tmp.q1TOnBC06b
Jan 29 00:53:45 + file_diff_from_base /tmp/tmp.q1TOnBC06b

pytorch_linux_xenial_cuda9_2_cudnn7_py3_gcc5_4_build (7/9)

Step: "(Optional) Merge target branch" (full log | diagnosis details | 🔁 rerun)

Automatic merge failed; fix conflicts and then commit the result.

CONFLICT (add/add): Merge conflict in .jenkins/caffe2/test.sh
Auto-merging .jenkins/caffe2/test.sh
CONFLICT (add/add): Merge conflict in .gitmodules
Auto-merging .gitmodules
CONFLICT (add/add): Merge conflict in .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
Auto-merging .circleci/verbatim-sources/job-specs/pytorch-job-specs.yml
CONFLICT (add/add): Merge conflict in .circleci/scripts/python_doc_push_script.sh
Auto-merging .circleci/scripts/python_doc_push_script.sh
CONFLICT (add/add): Merge conflict in .circleci/config.yml
Auto-merging .circleci/config.yml
Automatic merge failed; fix conflicts and then commit the result.


Exited with code exit status 1

pytorch_windows_vs2019_py36_cuda10.1_test1 (8/9)

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

RuntimeError: distributed/test_c10d failed!

  File "C:\Users\circleci\project\build\win_tmp\build\torch\distributed\algorithms\ddp_comm_hooks\__init__.py", line 7, in <module>
    from . import (
  File "C:\Users\circleci\project\build\win_tmp\build\torch\distributed\algorithms\ddp_comm_hooks\powerSGD_hook.py", line 7, in <module>
    import torch.distributed.algorithms.ddp_comm_hooks.default_hooks as default
AttributeError: module 'torch.distributed.algorithms' has no attribute 'ddp_comm_hooks'
Traceback (most recent call last):
  File "run_test.py", line 922, in <module>
    main()
  File "run_test.py", line 901, in main
    raise RuntimeError(err_message)
RuntimeError: distributed/test_c10d failed!

(base) C:\Users\circleci\project\test>if ERRORLEVEL 1 exit /b 1 
+ cleanup
+ retcode=1
+ set +x


Exited with code exit status 1

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test1 (9/9)

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

Jan 29 01:12:12 sccache: error: couldn't connect to server

Jan 29 01:12:12 +++ eval 'extract_trap_cmd '
Jan 29 01:12:12 ++++ extract_trap_cmd
Jan 29 01:12:12 ++++ printf '%s\n' ''
Jan 29 01:12:12 +++ printf '%s\n' cleanup
Jan 29 01:12:12 ++ trap -- '
Jan 29 01:12:12 cleanup' EXIT
Jan 29 01:12:12 ++ [[ pytorch-linux-xenial-cuda10.2-cudnn7-py3-gcc7-test1 != *pytorch-win-* ]]
Jan 29 01:12:12 ++ which sccache
Jan 29 01:12:12 ++ sccache --stop-server
Jan 29 01:12:12 Stopping sccache server...
Jan 29 01:12:12 sccache: error: couldn't connect to server
Jan 29 01:12:12 sccache: caused by: Connection refused (os error 111)
Jan 29 01:12:12 ++ true
Jan 29 01:12:12 ++ rm /var/lib/jenkins/sccache_error.log
Jan 29 01:12:12 ++ [[ pytorch-linux-xenial-cuda10.2-cudnn7-py3-gcc7-test1 == *rocm* ]]
Jan 29 01:12:12 ++ SCCACHE_ERROR_LOG=/var/lib/jenkins/sccache_error.log
Jan 29 01:12:12 ++ SCCACHE_IDLE_TIMEOUT=1200
Jan 29 01:12:12 ++ RUST_LOG=sccache::server=error
Jan 29 01:12:12 ++ sccache --start-server
Jan 29 01:12:12 sccache: Starting the server...
Jan 29 01:12:12 ++ sccache --zero-stats

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 to the (internal) Dr. CI Users group.

…SGD_hook.py by creating a util function that make a vanilla allreduce future" Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26070147](https://our.internmc.facebook.com/intern/diff/D26070147/) [ghstack-poisoned]

… by creating a util function that make a vanilla allreduce future Pull Request resolved: #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 ghstack-source-id: 120376248 Differential Revision: [D26070147](https://our.internmc.facebook.com/intern/diff/D26070147/)

…SGD_hook.py by creating a util function that make a vanilla allreduce future" Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26070147](https://our.internmc.facebook.com/intern/diff/D26070147/) [ghstack-poisoned]

… by creating a util function that make a vanilla allreduce future Pull Request resolved: #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 ghstack-source-id: 120619680 Differential Revision: [D26070147](https://our.internmc.facebook.com/intern/diff/D26070147/)

facebook-github-bot · 2021-01-29T03:03:42Z

This pull request has been merged in e7b3496.

facebook-github-bot · 2021-01-29T17:10:09Z

This pull request has been reverted by 5a406c0.

izdeby · 2021-01-29T17:11:15Z

Reverting due to a broken CI

rohan-varma · 2021-01-29T18:23:27Z

@SciPioneer Looks like the failures on this PR were legit:

Jan 29 01:19:21 Traceback (most recent call last):
Jan 29 01:19:21   File "distributed/test_c10d.py", line 21, in <module>
Jan 29 01:19:21     import torch.distributed.algorithms.ddp_comm_hooks.default_hooks as default
Jan 29 01:19:21   File "/opt/conda/lib/python3.6/site-packages/torch/distributed/algorithms/ddp_comm_hooks/__init__.py", line 7, in <module>
Jan 29 01:19:21     from . import (
Jan 29 01:19:21   File "/opt/conda/lib/python3.6/site-packages/torch/distributed/algorithms/ddp_comm_hooks/powerSGD_hook.py", line 7, in <module>
Jan 29 01:19:21     import torch.distributed.algorithms.ddp_comm_hooks.default_hooks as default
Jan 29 01:19:21 AttributeError: module 'torch.distributed.algorithms' has no attribute 'ddp_comm_hooks'
Jan 29 01:19:21 Traceback (most recent call last):
Jan 29 01:19:21   File "test/run_test.py", line 922, in <module>
Jan 29 01:19:21     main()
Jan 29 01:19:21   File "test/run_test.py", line 901, in main
Jan 29 01:19:21     raise RuntimeError(err_message)
Jan 29 01:19:21 RuntimeError: distributed/test_c10d failed!
Jan 29 01:19:22 + cleanup
Jan 29 01:19:22 + retcode=1
Jan 29 01:19:22 + set +x

…werSGD_hook.py by creating a util function that make a vanilla allreduce future Resubmission of #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26162333](https://our.internmc.facebook.com/intern/diff/D26162333/) [ghstack-poisoned]

…s.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future" Resubmission of #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26162333](https://our.internmc.facebook.com/intern/diff/D26162333/) [ghstack-poisoned]

…werSGD_hook.py by creating a util function that make a vanilla allreduce future Pull Request resolved: #51400 Resubmission of #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 ghstack-source-id: 120715333 Differential Revision: [D26162333](https://our.internmc.facebook.com/intern/diff/D26162333/)

…s.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future" Resubmission of #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 Differential Revision: [D26162333](https://our.internmc.facebook.com/intern/diff/D26162333/) [ghstack-poisoned]

…werSGD_hook.py by creating a util function that make a vanilla allreduce future (#51400) Summary: Pull Request resolved: #51400 Resubmission of #51094 Address #50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression #47202 ghstack-source-id: 120725690 Test Plan: buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_powerSGD_ddp_comm_hook_nccl buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_default_ddp_comm_hooks_nccl Reviewed By: rohan-varma Differential Revision: D26162333 fbshipit-source-id: ccc2eae5383a23673e00d61cb5570fb8bf749cd0

… by creating a util function that make a vanilla allreduce future (pytorch#51094) Summary: Pull Request resolved: pytorch#51094 Address pytorch#50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression pytorch#47202 ghstack-source-id: 120619680 Test Plan: buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_powerSGD_ddp_comm_hook_nccl buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_default_ddp_comm_hooks_nccl Reviewed By: rohan-varma Differential Revision: D26070147 fbshipit-source-id: 8c9339f1511e8f24cc906b9411cfe4850a5a6d81

…werSGD_hook.py by creating a util function that make a vanilla allreduce future (pytorch#51400) Summary: Pull Request resolved: pytorch#51400 Resubmission of pytorch#51094 Address pytorch#50973 (comment) Original PR issue: Investigate Applying PowerSGD to Communication Hook for Gradient Compression pytorch#47202 ghstack-source-id: 120725690 Test Plan: buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_powerSGD_ddp_comm_hook_nccl buck test mode/dev-nosan caffe2/test/distributed:c10d -- test_default_ddp_comm_hooks_nccl Reviewed By: rohan-varma Differential Revision: D26162333 fbshipit-source-id: ccc2eae5383a23673e00d61cb5570fb8bf749cd0

wayi1 requested review from mingzhe09088, mrshenli, pritamdamania87, rohan-varma and zhaojuanmao as code owners January 26, 2021 06:34

facebook-github-bot added cla signed oncall: distributed Add this issue/PR to distributed oncall triage queue labels Jan 26, 2021

wayi1 mentioned this pull request Jan 28, 2021

[Gradient Compression] Allow BatchedPowerSGD to run vanilla allreduce for the first K iterations #51270

Closed

rohan-varma reviewed Jan 28, 2021

View reviewed changes

Comment thread torch/distributed/algorithms/ddp_comm_hooks/default_hooks.py

rohan-varma reviewed Jan 28, 2021

View reviewed changes

Comment thread torch/distributed/algorithms/ddp_comm_hooks/default_hooks.py Outdated

rohan-varma approved these changes Jan 28, 2021

View reviewed changes

wayi added 2 commits January 28, 2021 16:02

facebook-github-bot closed this in e7b3496 Jan 29, 2021

facebook-github-bot added the Merged label Jan 29, 2021

facebook-github-bot added the Reverted label Jan 29, 2021

wayi1 mentioned this pull request Jan 30, 2021

[Resubmission][Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future #51400

Closed

facebook-github-bot deleted the gh/SciPioneer/49/head branch February 1, 2021 15:19

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future#51094

[Gradient Compression] Refactor default_hooks.py and powerSGD_hook.py by creating a util function that make a vanilla allreduce future#51094
wayi1 wants to merge 4 commits intogh/SciPioneer/49/basefrom
gh/SciPioneer/49/head

wayi1 commented Jan 26, 2021 •

edited

Loading

Uh oh!

facebook-github-bot commented Jan 26, 2021 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

facebook-github-bot commented Jan 29, 2021

Uh oh!

facebook-github-bot commented Jan 29, 2021

Uh oh!

izdeby commented Jan 29, 2021

Uh oh!

rohan-varma commented Jan 29, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

wayi1 commented Jan 26, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

facebook-github-bot commented Jan 26, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🕵️ 9 new failures recognized by patterns

pytorch_linux_xenial_py3_clang5_asan_test2 (1/9)

pytorch_linux_bionic_py3_6_clang9_test (2/9)

pytorch_xla_linux_bionic_py3_6_clang9_build (3/9)

pytorch_linux_xenial_py3_6_gcc5_4_build (4/9)

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test2 (5/9)

pytorch_linux_xenial_py3_clang5_asan_test1 (6/9)

pytorch_linux_xenial_cuda9_2_cudnn7_py3_gcc5_4_build (7/9)

pytorch_windows_vs2019_py36_cuda10.1_test1 (8/9)

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test1 (9/9)

Uh oh!

Uh oh!

Uh oh!

facebook-github-bot commented Jan 29, 2021

Uh oh!

facebook-github-bot commented Jan 29, 2021

Uh oh!

izdeby commented Jan 29, 2021

Uh oh!

rohan-varma commented Jan 29, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

wayi1 commented Jan 26, 2021 •

edited

Loading

facebook-github-bot commented Jan 26, 2021 •

edited

Loading