[Cutlass 3.2.2 submodule upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2 #112762

kadeng · 2023-11-02T18:11:56Z

Stack from ghstack (oldest at bottom):

The inductor cutlass backend was written against Cutlass version 3.1.x,
there are some incompatible changes in Cutlass 3.2.2 which the
Inductor cutlass backend needs to adapt to.

Test plan:

If third_party/cutlass is upgraded to Cutlass tag v3.2.2,
several tests within test/inductor/test_max_autotune.py start to
fail. With this diff applied, they pass again.

cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @muchulee8 @aakhundov @ColinPeppler

Differential Revision: D50986555

The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With thid diff, they pass again. [ghstack-poisoned]

pytorch-bot · 2023-11-02T18:11:59Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/112762

📄 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! (1 Unrelated Failure)

As of commit 2705932 with merge base dc1a358 ():

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

trunk / macos-12-py3-arm64 / test (default, 3, 3, macos-m1-12, unstable) (gh)

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

aakhundov · 2023-11-02T18:21:00Z

Thanks @kadeng! Do we need to enable ciflow/trunk, too?

kadeng · 2023-11-02T21:39:56Z

Currently a Windows build failure is blocking this. A Cutlass Issue has been created here: NVIDIA/cutlass#1171

…nductor cutlass backend to Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler [ghstack-poisoned]

…ade] Adapt Inductor cutlass backend to Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler [ghstack-poisoned]

…ass 3.2.2 upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler [ghstack-poisoned]

…Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler [ghstack-poisoned]

kadeng · 2023-11-03T14:14:28Z

CMakeLists.txt

 endif()

 if(MSVC)
+  # MSVC by default does not apply the correct __cplusplus version as specified by the C++ standard


Resolving Windows build issue, see NVIDIA/cutlass#1171 (comment)

If this does fix the problem I am not sure if we want to set this global, @malfet will likely have thoughts on this.

That being said it is possible to set flags on specific targets, see:

pytorch/caffe2/CMakeLists.txt

Lines 1820 to 1832 in d084a02

if(MSVC)

# This is used to enable the conforming lambda processor in MSVC

# Which allows us to capture constexpr in lambdas

# Note that this will be turned on by default for std=c++20 and above

# This should be applied globally when https://github.com/pytorch/pytorch/issues/92600 is fixed

foreach(tmp ${MEM_EFF_ATTENTION_CUDA_SOURCES})

# MEM_EFF_ATTENTION_CUDA is populated in pytorch/aten/src/ATen/CMakeLists.txt

# We iterate over these files, updating paths and adding the compile flag

FILE(RELATIVE_PATH tmp_path "${PROJECT_SOURCE_DIR}" "${tmp}")

SET(tmp_path "../${tmp_path}")

set_source_files_properties(${tmp_path} PROPERTIES COMPILE_FLAGS "-Xcompiler /Zc:lambda")

endforeach()

endif()

As mentioned in the chat, I think we want to set this globally, since it restores MSVC to C++-Standard compliant behavior.

See

C++ Standard

MSVC Docs which pretty clearly states that the default behavior of MSVC is a hack.

ipiszy

Thanks @kadeng !

…Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler [ghstack-poisoned]

kadeng · 2023-11-03T20:05:02Z

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

kadeng · 2023-11-03T20:10:10Z

@pytorchbot merge

pytorchmergebot · 2023-11-03T20:12:53Z

This PR updates submodules third_party/cutlass

If those updates are intentional, please add "submodule" keyword to PR title/description.

kadeng · 2023-11-03T20:14:32Z

@pytorchbot merge

kadeng · 2023-11-03T20:18:32Z

Triggered re-run of all checks...

…e upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler Differential Revision: [D50986555](https://our.internmc.facebook.com/intern/diff/D50986555) [ghstack-poisoned]

kadeng · 2023-11-03T20:27:46Z

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

kadeng · 2023-11-03T20:35:02Z

@pytorchbot merge

pytorchmergebot · 2023-11-03T20:36:48Z

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

…ackend to Cutlass 3.2.2" The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 muchulee8 aakhundov ColinPeppler Differential Revision: [D50986555](https://our.internmc.facebook.com/intern/diff/D50986555) [ghstack-poisoned]

pytorchmergebot · 2023-11-03T20:42:02Z

Merge failed

Reason: New commits were pushed while merging. Please rerun the merge command.

Details for Dev Infra team

Raised by workflow job

kadeng · 2023-11-03T20:42:30Z

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

kadeng · 2023-11-03T20:43:54Z

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

kadeng · 2023-11-03T20:45:29Z

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

kadeng · 2023-11-03T20:48:18Z

@pytorchbot merge

pytorchmergebot · 2023-11-03T20:50:08Z

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

Summary: This PR adds epilogue fusion code generation support for the new experimental [Inductor Cutlass backend]([#108015]). Details: A fusion happens on the GEMM template level by taking a Cutlass 3.x GEMM Universal Matmul Kernel template and adding a custom template functor based on Cutlass new “Epilogue Visitor Trees” (EVT) on top, which represents and performs the computation of the fused Pointwise / Elementwise computation nodes. This is the approach dictated by [NVIDIA/cutlass example 49](https://github.com/NVIDIA/cutlass/blob/main/examples/49_hopper_gemm_with_collective_builder/49_collective_builder.cu), which is currently the only documentation and example of Cutlass Epilogue Visitor Trees. This EVT functor in turn is a hierarchical template expression which represents an abstract syntax tree of the fused computation to perform. A second codegen task is to create a hierarchical initializer expression, which provides potentially necessary arguments to each of the functor subexpressions. Step 1 functionality: * End to end code generation is possible using the above approach. * Supports simple elementwise expression fusion of chains of elementwise operations (with scalar constants ) after a matmul. * Elementwise operation support includes addition, subtraction, multiplication, division, minimum, maximum etc. * Examples / Unit tests include ReLU and ReLU6 fusion. * Support for fp16 and fp16 with fp32 accumulation data types. * Generates SM90 ( Hopper ) based CUDA Kernels ( as Cutlass up to 3.2.0 only supported EVT for SM90 ) The following is not yet supported, and is left for future work: * Full operation support ( e.g. full set of all ops usually handled via V.ops handlers ) * Cutlass EVT with SM80 support ( possible in Cutlass 3.2.1 according to release notes, but not yet documented ) * Add support for additional (auxiliary) inputs, which changes the Template Kernels' call signature * Add support for additional (auxiliary) outputs ( requires support for full computation graphs ) * Add support for reduction operations and operations which use different output layouts than the input * Add support for additional dtypes ( as far as Cutlass allows ) This PR updates third_party/cutlass to v3.2.2, which has some important improvements and features for the inductor backend. See also Cutlass release notes: https://github.com/NVIDIA/cutlass/releases/tag/v3.2.1 and https://github.com/NVIDIA/cutlass/releases/tag/v3.2.2 Notable changes in Cutlass 3.2.1 include: * Cutlass codegen python code has moved into a package with the "cutlass_library" namespace, which allows to prevent namespace clashes without resolving to monkey-patching ( which was done earlier ). * Support for SM80 epilogue visitor trees ( according to the Release Notes, not tried yet ) * Small API changes to the cutlass_library API ( requires adapting the inductor backend code ) Notable changes in Cutlass 3.2.2 include: * Bugfix that led to CUDA Illegal memory access in some Pytorch unit tests involving flash attention Test Plan: * CI * pytest test/inductor/test_max_autotune.py Note: So far, the CUTLASS backend is still disabled by default. Benchmarks are planned once more advanced fusions are enabled. Differential Revision: [D50988161](https://our.internmc.facebook.com/intern/diff/D50988161) Pull Request resolved: #110890 Approved by: https://github.com/jansel ghstack dependencies: #112762

…utlass 3.2.2 (pytorch#112762) The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. Differential Revision: [D50986555](https://our.internmc.facebook.com/intern/diff/D50986555) Pull Request resolved: pytorch#112762 Approved by: https://github.com/ipiszy, https://github.com/drisspg

…110890) Summary: This PR adds epilogue fusion code generation support for the new experimental [Inductor Cutlass backend]([pytorch#108015]). Details: A fusion happens on the GEMM template level by taking a Cutlass 3.x GEMM Universal Matmul Kernel template and adding a custom template functor based on Cutlass new “Epilogue Visitor Trees” (EVT) on top, which represents and performs the computation of the fused Pointwise / Elementwise computation nodes. This is the approach dictated by [NVIDIA/cutlass example 49](https://github.com/NVIDIA/cutlass/blob/main/examples/49_hopper_gemm_with_collective_builder/49_collective_builder.cu), which is currently the only documentation and example of Cutlass Epilogue Visitor Trees. This EVT functor in turn is a hierarchical template expression which represents an abstract syntax tree of the fused computation to perform. A second codegen task is to create a hierarchical initializer expression, which provides potentially necessary arguments to each of the functor subexpressions. Step 1 functionality: * End to end code generation is possible using the above approach. * Supports simple elementwise expression fusion of chains of elementwise operations (with scalar constants ) after a matmul. * Elementwise operation support includes addition, subtraction, multiplication, division, minimum, maximum etc. * Examples / Unit tests include ReLU and ReLU6 fusion. * Support for fp16 and fp16 with fp32 accumulation data types. * Generates SM90 ( Hopper ) based CUDA Kernels ( as Cutlass up to 3.2.0 only supported EVT for SM90 ) The following is not yet supported, and is left for future work: * Full operation support ( e.g. full set of all ops usually handled via V.ops handlers ) * Cutlass EVT with SM80 support ( possible in Cutlass 3.2.1 according to release notes, but not yet documented ) * Add support for additional (auxiliary) inputs, which changes the Template Kernels' call signature * Add support for additional (auxiliary) outputs ( requires support for full computation graphs ) * Add support for reduction operations and operations which use different output layouts than the input * Add support for additional dtypes ( as far as Cutlass allows ) This PR updates third_party/cutlass to v3.2.2, which has some important improvements and features for the inductor backend. See also Cutlass release notes: https://github.com/NVIDIA/cutlass/releases/tag/v3.2.1 and https://github.com/NVIDIA/cutlass/releases/tag/v3.2.2 Notable changes in Cutlass 3.2.1 include: * Cutlass codegen python code has moved into a package with the "cutlass_library" namespace, which allows to prevent namespace clashes without resolving to monkey-patching ( which was done earlier ). * Support for SM80 epilogue visitor trees ( according to the Release Notes, not tried yet ) * Small API changes to the cutlass_library API ( requires adapting the inductor backend code ) Notable changes in Cutlass 3.2.2 include: * Bugfix that led to CUDA Illegal memory access in some Pytorch unit tests involving flash attention Test Plan: * CI * pytest test/inductor/test_max_autotune.py Note: So far, the CUTLASS backend is still disabled by default. Benchmarks are planned once more advanced fusions are enabled. Differential Revision: [D50988161](https://our.internmc.facebook.com/intern/diff/D50988161) Pull Request resolved: pytorch#110890 Approved by: https://github.com/jansel ghstack dependencies: pytorch#112762

…utlass 3.2.2 (pytorch#112762) The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With this diff applied, they pass again. Differential Revision: [D50986555](https://our.internmc.facebook.com/intern/diff/D50986555) Pull Request resolved: pytorch#112762 Approved by: https://github.com/ipiszy, https://github.com/drisspg

…110890) Summary: This PR adds epilogue fusion code generation support for the new experimental [Inductor Cutlass backend]([pytorch#108015]). Details: A fusion happens on the GEMM template level by taking a Cutlass 3.x GEMM Universal Matmul Kernel template and adding a custom template functor based on Cutlass new “Epilogue Visitor Trees” (EVT) on top, which represents and performs the computation of the fused Pointwise / Elementwise computation nodes. This is the approach dictated by [NVIDIA/cutlass example 49](https://github.com/NVIDIA/cutlass/blob/main/examples/49_hopper_gemm_with_collective_builder/49_collective_builder.cu), which is currently the only documentation and example of Cutlass Epilogue Visitor Trees. This EVT functor in turn is a hierarchical template expression which represents an abstract syntax tree of the fused computation to perform. A second codegen task is to create a hierarchical initializer expression, which provides potentially necessary arguments to each of the functor subexpressions. Step 1 functionality: * End to end code generation is possible using the above approach. * Supports simple elementwise expression fusion of chains of elementwise operations (with scalar constants ) after a matmul. * Elementwise operation support includes addition, subtraction, multiplication, division, minimum, maximum etc. * Examples / Unit tests include ReLU and ReLU6 fusion. * Support for fp16 and fp16 with fp32 accumulation data types. * Generates SM90 ( Hopper ) based CUDA Kernels ( as Cutlass up to 3.2.0 only supported EVT for SM90 ) The following is not yet supported, and is left for future work: * Full operation support ( e.g. full set of all ops usually handled via V.ops handlers ) * Cutlass EVT with SM80 support ( possible in Cutlass 3.2.1 according to release notes, but not yet documented ) * Add support for additional (auxiliary) inputs, which changes the Template Kernels' call signature * Add support for additional (auxiliary) outputs ( requires support for full computation graphs ) * Add support for reduction operations and operations which use different output layouts than the input * Add support for additional dtypes ( as far as Cutlass allows ) This PR updates third_party/cutlass to v3.2.2, which has some important improvements and features for the inductor backend. See also Cutlass release notes: https://github.com/NVIDIA/cutlass/releases/tag/v3.2.1 and https://github.com/NVIDIA/cutlass/releases/tag/v3.2.2 Notable changes in Cutlass 3.2.1 include: * Cutlass codegen python code has moved into a package with the "cutlass_library" namespace, which allows to prevent namespace clashes without resolving to monkey-patching ( which was done earlier ). * Support for SM80 epilogue visitor trees ( according to the Release Notes, not tried yet ) * Small API changes to the cutlass_library API ( requires adapting the inductor backend code ) Notable changes in Cutlass 3.2.2 include: * Bugfix that led to CUDA Illegal memory access in some Pytorch unit tests involving flash attention Test Plan: * CI * pytest test/inductor/test_max_autotune.py Note: So far, the CUTLASS backend is still disabled by default. Benchmarks are planned once more advanced fusions are enabled. Differential Revision: [D50988161](https://our.internmc.facebook.com/intern/diff/D50988161) Pull Request resolved: pytorch#110890 Approved by: https://github.com/jansel ghstack dependencies: pytorch#112762

The inductor cutlass backend was written against Cutlass version 3.1.x, there are some incompatible changes in Cutlass 3.2.2 which the Inductor cutlass backend needs to adapt to. Test plan: If third_party/cutlass is upgraded to Cutlass tag v3.2.2, several tests within test/inductor/test_max_autotune.py start to fail. With thid diff, they pass again. ghstack-source-id: d2f98d9 Pull Request resolved: pytorch/pytorch#112762

github-actions bot added module: inductor ciflow/inductor labels Nov 2, 2023

kadeng requested a review from drisspg November 2, 2023 18:12

kadeng mentioned this pull request Nov 2, 2023

[Inductor CUTLASS backend] Epilogue fusion codegen (Step 1) #110890

Closed

kadeng added ciflow/trunk Trigger trunk jobs on your pull request topic: not user facing topic category labels Nov 2, 2023

kadeng mentioned this pull request Nov 2, 2023

Build issue on Windows / VSCode 2019 ( blocking Pytorch to upgrade Cutlass ) NVIDIA/cutlass#1171

Closed

kadeng added 3 commits November 3, 2023 04:03

kadeng added 2 commits November 3, 2023 07:05

kadeng commented Nov 3, 2023

View reviewed changes

kadeng mentioned this pull request Nov 3, 2023

[Cutlass 3.3 submodule upgrade] #112861

Closed

kadeng marked this pull request as ready for review November 3, 2023 17:45

ipiszy approved these changes Nov 3, 2023

View reviewed changes

drisspg approved these changes Nov 3, 2023

View reviewed changes

kadeng added 2 commits November 3, 2023 11:23

kadeng changed the title ~~[Cutlass 3.2.2 upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2~~ [Cutlass 3.2.2 submodule upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2 Nov 3, 2023

pytorchmergebot removed the merging label Nov 3, 2023

pytorch deleted a comment from pytorchmergebot Nov 3, 2023

pytorchmergebot added the merging label Nov 3, 2023

pytorchmergebot removed the merging label Nov 3, 2023

pytorchmergebot added the merging label Nov 3, 2023

pytorchmergebot added Merged and removed merging labels Nov 4, 2023

pytorchmergebot closed this in e36dba3 Nov 4, 2023

kadeng mentioned this pull request Nov 6, 2023

Update to newest CUTLASS version 3.2.2 #108070

Closed

facebook-github-bot deleted the gh/kadeng/5/head branch November 7, 2023 15:24

	if(MSVC)
	# This is used to enable the conforming lambda processor in MSVC
	# Which allows us to capture constexpr in lambdas
	# Note that this will be turned on by default for std=c++20 and above
	# This should be applied globally when https://github.com/pytorch/pytorch/issues/92600 is fixed
	foreach(tmp ${MEM_EFF_ATTENTION_CUDA_SOURCES})
	# MEM_EFF_ATTENTION_CUDA is populated in pytorch/aten/src/ATen/CMakeLists.txt
	# We iterate over these files, updating paths and adding the compile flag
	FILE(RELATIVE_PATH tmp_path "${PROJECT_SOURCE_DIR}" "${tmp}")
	SET(tmp_path "../${tmp_path}")
	set_source_files_properties(${tmp_path} PROPERTIES COMPILE_FLAGS "-Xcompiler /Zc:lambda")
	endforeach()
	endif()

[Cutlass 3.2.2 submodule upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2 #112762

[Cutlass 3.2.2 submodule upgrade] Adapt Inductor cutlass backend to Cutlass 3.2.2 #112762

Uh oh!

Conversation

kadeng commented Nov 2, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented Nov 2, 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/112762

✅ You can merge normally! (1 Unrelated Failure)

Uh oh!

aakhundov commented Nov 2, 2023

Uh oh!

kadeng commented Nov 2, 2023

Uh oh!

kadeng Nov 3, 2023

Choose a reason for hiding this comment

Uh oh!

drisspg Nov 3, 2023

Choose a reason for hiding this comment

Uh oh!

kadeng Nov 3, 2023

Choose a reason for hiding this comment

Uh oh!

ipiszy left a comment

Choose a reason for hiding this comment

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

pytorchmergebot commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

pytorchmergebot commented Nov 3, 2023

Merge started

Uh oh!

pytorchmergebot commented Nov 3, 2023

Merge failed

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

kadeng commented Nov 3, 2023

Uh oh!

pytorchmergebot commented Nov 3, 2023

Merge started

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

kadeng commented Nov 2, 2023 •

edited

Loading

pytorch-bot bot commented Nov 2, 2023 •

edited

Loading