Change aot_compile callsites #2207
Closed
+14
−1
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
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 25, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
f2d7def to
fc343b4
Compare
angelayi
added a commit
to angelayi/pytorch
that referenced
this pull request
Mar 25, 2024
Summary: X-link: pytorch/benchmark#2207 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Test Plan: CI Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
angelayi
added a commit
to angelayi/pytorch
that referenced
this pull request
Mar 25, 2024
Summary: X-link: pytorch/benchmark#2207 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Test Plan: CI Reviewed By: SherlockNoMad Differential Revision: D54808612
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
fc343b4 to
871039e
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
871039e to
1c4bab9
Compare
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
1c4bab9 to
3e160db
Compare
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
39599b5 to
6752f30
Compare
angelayi
added a commit
to angelayi/pytorch
that referenced
this pull request
Mar 26, 2024
Summary: X-link: pytorch/benchmark#2207 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Test Plan: CI Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
6752f30 to
c117a80
Compare
angelayi
added a commit
to angelayi/benchmark
that referenced
this pull request
Mar 28, 2024
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
Summary: X-link: pytorch/pytorch#122225 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Reviewed By: SherlockNoMad Differential Revision: D54808612
c117a80 to
1b1d20f
Compare
angelayi
added a commit
to angelayi/pytorch
that referenced
this pull request
Mar 29, 2024
Summary: X-link: pytorch/benchmark#2207 Replacing `torch._export.aot_compile` callsites with ``` ep = torch.export._export(.., predispatch=True) # Traces the given program into predispatch IR so_path = torch._inductor.aot_compile_ep(ep, ...) # Takes an exported program and compiles it into a .so ``` This allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do `export + serialize + deserialize + inductor` to mimic internal production use cases better. This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know. This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site. Test Plan: CI Reviewed By: SherlockNoMad Differential Revision: D54808612
|
This pull request was exported from Phabricator. Differential Revision: D54808612 |
|
This pull request has been merged in 756ea35. |
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.
Summary:
X-link: pytorch/pytorch#122225
Replacing
torch._export.aot_compilecallsites withThis allows us to explicitly split up the export step from AOTInductor. We can later modify tests to do
export + serialize + deserialize + inductorto mimic internal production use cases better.This PR also enables export's predispatch IR most of the the existing use cases. Previously this is using export to torch IR, which is a different graph. This may result in some performance regressions as some of inductor's passes will no longer run -- if so, please let me know.
This PR changes the seemingly low-impact files to the new calling convention, and a followup PR will change the high-impact site.
Reviewed By: SherlockNoMad
Differential Revision: D54808612