[inductor][cpp] epilogue support for gemm template #126019
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/126019
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 10174bd with merge base cb6ef68 ( UNSTABLE - The following job failed but was 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. |
jgong5
added a commit
that referenced
this pull request
May 12, 2024
ghstack-source-id: 5c5aa78127c399dc804cc7f768fe038cbf05a7e4 Pull Request resolved: #126019
jgong5
added a commit
that referenced
this pull request
May 13, 2024
ghstack-source-id: 58c56a7ef3271a127573415e5391b8f1ac5d1875 Pull Request resolved: #126019
This was referenced May 13, 2024
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new `codegen_loop_bodies` and `codegen_functions` methods are added to c++ vector codegen for this purpose. This is leveraged by the `store_output` method of the template kernel for epilogue codegen and store to the final result. cc voznesenskym penguinwu EikanWang Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new `codegen_loop_bodies` and `codegen_functions` methods are added to c++ vector codegen for this purpose. This is leveraged by the `store_output` method of the template kernel for epilogue codegen and store to the final result. cc voznesenskym penguinwu EikanWang Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new `codegen_loop_bodies` and `codegen_functions` methods are added to c++ vector codegen for this purpose. This is leveraged by the `store_output` method of the template kernel for epilogue codegen and store to the final result. cc voznesenskym penguinwu EikanWang Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
jansel
approved these changes
May 15, 2024
|
@pytorchbot merge |
Merge startedYour 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 |
pytorchmergebot
pushed a commit
that referenced
this pull request
May 16, 2024
…ue fusion (#126068) As part of #125683, this PR adds the initial bf16/fp16 gemm template support with micro-gemm implemented with fused type casting and fp32 computation. It doesn't provide epilogue fusion support yet which will be added in the next PR. Pull Request resolved: #126068 Approved by: https://github.com/jansel ghstack dependencies: #126019
pytorchmergebot
pushed a commit
that referenced
this pull request
May 23, 2024
…ue fusion (#126068) As part of #125683, this PR adds the initial bf16/fp16 gemm template support with micro-gemm implemented with fused type casting and fp32 computation. It doesn't provide epilogue fusion support yet which will be added in the next PR. Pull Request resolved: #126068 Approved by: https://github.com/jansel ghstack dependencies: #124021, #126019
pytorchmergebot
added a commit
that referenced
this pull request
May 23, 2024
This reverts commit 57108d9. Reverted #126019 on behalf of https://github.com/huydhn due to Sorry for reverting your change but I think it has a land race and failing in trunk https://hud.pytorch.org/pytorch/pytorch/commit/2ac33a9f663269e6060246337c776a20c3b7c858 ([comment](#124021 (comment)))
|
@jgong5 your PR has been successfully reverted. |
leslie-fang-intel
pushed a commit
to leslie-fang-intel/pytorch
that referenced
this pull request
May 23, 2024
ghstack-source-id: 6971d95cd35d2c5a399f851848e8b2250176df52 Pull Request resolved: pytorch#126019
pytorchmergebot
pushed a commit
that referenced
this pull request
May 23, 2024
…ue fusion (#126068) As part of #125683, this PR adds the initial bf16/fp16 gemm template support with micro-gemm implemented with fused type casting and fp32 computation. It doesn't provide epilogue fusion support yet which will be added in the next PR. Pull Request resolved: #126068 Approved by: https://github.com/jansel ghstack dependencies: #124021, #126019
pytorchmergebot
pushed a commit
that referenced
this pull request
May 23, 2024
) As part of #125683, this PR adds epilogue fusion support for bf16/fp16 gemms. The key changes are as follows: 1. bf16 linear w/ epilogue fusion of some ops was originally supported via ATen oneDNN linear pointwise ops. In order to match the ATen op semantics, in-template epilogue support is added to the cpp gemm template so that we would have: "gemm + in-template epilogues -> template buffer". If the template is chosen for codegen, the in-template epilogues will be concatenated with the out-of-template epilogues that are appended during the scheduling. 2. Support bf16/fp16 legalization for `codegen_loop_bodies` which is used to generate the epilogue loops. 3. We used to leverage the in-place buffer mechanism to handle the in-place buffers in the epilogue codegen, in particular, for the reuses for output buffers of GEMM, template and epilogues. This is not correct since the output buffer is an "output" not an "in-place" buffer of the template kernel itself. Now, we use a dedicated "aliases" dict to manage such buffer reuses and the intermediate aliasing buffers are removed after codegen. 4. Add `localize_buffer` method to `LocalBufferScope` to allow the replacement of a global buffer with a local one in the given inductor IR nodes. This helps the fused loops to work on smaller-sized local buffers for better data locality. Pull Request resolved: #126545 Approved by: https://github.com/jansel ghstack dependencies: #124021, #126019, #126068
pytorchmergebot
added a commit
that referenced
this pull request
May 23, 2024
This reverts commit 08f57b4. Reverted #126019 on behalf of https://github.com/DanilBaibak due to Broken trunk ([comment](#124021 (comment)))
|
@jgong5 your PR has been successfully reverted. |
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new `codegen_loop_bodies` and `codegen_functions` methods are added to c++ vector codegen for this purpose. This is leveraged by the `store_output` method of the template kernel for epilogue codegen and store to the final result. cc voznesenskym penguinwu EikanWang Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new `codegen_loop_bodies` and `codegen_functions` methods are added to c++ vector codegen for this purpose. This is leveraged by the `store_output` method of the template kernel for epilogue codegen and store to the final result. cc voznesenskym penguinwu EikanWang Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
|
@pytorchbot rebase |
|
@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here |
|
Successfully rebased |
pytorchmergebot
pushed a commit
that referenced
this pull request
May 24, 2024
ghstack-source-id: a6d9d63cad99b8fd3b3ed6daff89676953fe38ed Pull Request resolved: #126019
|
@pytorchbot merge |
Merge startedYour 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 |
pytorchmergebot
pushed a commit
that referenced
this pull request
May 24, 2024
…ue fusion (#126068) As part of #125683, this PR adds the initial bf16/fp16 gemm template support with micro-gemm implemented with fused type casting and fp32 computation. It doesn't provide epilogue fusion support yet which will be added in the next PR. Pull Request resolved: #126068 Approved by: https://github.com/jansel ghstack dependencies: #124021, #126019
pytorchmergebot
pushed a commit
that referenced
this pull request
May 24, 2024
) As part of #125683, this PR adds epilogue fusion support for bf16/fp16 gemms. The key changes are as follows: 1. bf16 linear w/ epilogue fusion of some ops was originally supported via ATen oneDNN linear pointwise ops. In order to match the ATen op semantics, in-template epilogue support is added to the cpp gemm template so that we would have: "gemm + in-template epilogues -> template buffer". If the template is chosen for codegen, the in-template epilogues will be concatenated with the out-of-template epilogues that are appended during the scheduling. 2. Support bf16/fp16 legalization for `codegen_loop_bodies` which is used to generate the epilogue loops. 3. We used to leverage the in-place buffer mechanism to handle the in-place buffers in the epilogue codegen, in particular, for the reuses for output buffers of GEMM, template and epilogues. This is not correct since the output buffer is an "output" not an "in-place" buffer of the template kernel itself. Now, we use a dedicated "aliases" dict to manage such buffer reuses and the intermediate aliasing buffers are removed after codegen. 4. Add `localize_buffer` method to `LocalBufferScope` to allow the replacement of a global buffer with a local one in the given inductor IR nodes. This helps the fused loops to work on smaller-sized local buffers for better data locality. Pull Request resolved: #126545 Approved by: https://github.com/jansel ghstack dependencies: #124021, #126019, #126068
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
As part of #125683, this PR adds the epilogue support for c++ gemm template by reusing the c++ vector codegen on sub-slices of tensors. This is implemented by retracing the epilogue IR nodes with new ranges and offsets. The new
codegen_loop_bodiesandcodegen_functionsmethods are added to c++ vector codegen for this purpose. This is leveraged by thestore_outputmethod of the template kernel for epilogue codegen and store to the final result.cc @voznesenskym @penguinwu @EikanWang @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang