Introduce aoti_call_delegate HOP #145630
Closed
Introduce aoti_call_delegate HOP #145630
Conversation
This file contains hidden or 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/145630
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (4 Unrelated Failures)As of commit 133a4e8 with merge base 5a527fa ( BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
78919e2 to
60a629c
Compare
SherlockNoMad
added a commit
to SherlockNoMad/pytorch
that referenced
this pull request
Jan 24, 2025
Summary: Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering. I introduce a new HOP to address this. The schema is following ``` aoti_call_delegate(lower_moduel: AOTInductorEPModule, original_gm: fx.GraphModule, weights: List[Tensor], inputs: List[Tensor]) ``` There are a few problems exposed by HOP - AOTI expects a FX graph with weights as getattr nodes, aka stateful graph. HOP expect graph_module arguments to be stateless. Export serializer also expect a stateless graph. Currently, to make AOTI happy, I am making `original_gm` stateful, and bypassing the serialization for `original_gm`. - As a result, the HOP is not re-traceable, as functionalization on stateful graph module argument will fail. Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test Reviewed By: zhxchen17 Differential Revision: D68359391
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
60a629c to
aae8b65
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
SherlockNoMad
added a commit
to SherlockNoMad/pytorch
that referenced
this pull request
Jan 27, 2025
Summary: Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering. I introduce a new HOP to address this. The schema is following ``` aoti_call_delegate(lower_moduel: AOTInductorEPModule, original_gm: fx.GraphModule, weights: List[Tensor], inputs: List[Tensor]) ``` There are a few problems exposed by HOP - AOTI expects a FX graph with weights as getattr nodes, aka stateful graph. HOP expect graph_module arguments to be stateless. Export serializer also expect a stateless graph. Currently, to make AOTI happy, I am making `original_gm` stateful, and bypassing the serialization for `original_gm`. - As a result, the HOP is not re-traceable, as functionalization on stateful graph module argument will fail. Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test Reviewed By: zhxchen17 Differential Revision: D68359391
aae8b65 to
4f8e4e0
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
SherlockNoMad
commented
Jan 27, 2025
4f8e4e0 to
769110e
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
7682b7a to
243c3e3
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
243c3e3 to
64d65ee
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
angelayi
reviewed
Jan 29, 2025
| AOTI_LOWERED_MODULE = "AOTInductorEPModule" | ||
|
|
||
|
|
||
| class AOTICallDelegate(HigherOrderOperator): |
There was a problem hiding this comment.
is it valuable to make this op more general? Like couldn't it also work for MTIA too?
64d65ee to
db69de9
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
SherlockNoMad
added a commit
to SherlockNoMad/pytorch
that referenced
this pull request
Jan 29, 2025
Summary: Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering. I introduce a new HOP to address this. The schema is following ``` aoti_call_delegate(lower_moduel: AOTInductorEPModule, original_gm: fx.GraphModule, weights: List[Tensor], inputs: List[Tensor]) ``` There are a few problems exposed by HOP - AOTI expects a FX graph with weights as getattr nodes, aka stateful graph. HOP expect graph_module arguments to be stateless. Export serializer also expect a stateless graph. Currently, to make AOTI happy, I am making `original_gm` stateful, and bypassing the serialization for `original_gm`. - As a result, the HOP is not re-traceable, as functionalization on stateful graph module argument will fail. Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test Reviewed By: zhxchen17 Differential Revision: D68359391
db69de9 to
65c6c2d
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
65c6c2d to
3014224
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
SherlockNoMad
added a commit
to SherlockNoMad/pytorch
that referenced
this pull request
Jan 30, 2025
Summary: Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering. I introduce a new HOP to address this. The schema is following ``` aoti_call_delegate(lower_moduel: AOTInductorEPModule, original_gm: fx.GraphModule, weights: List[Tensor], inputs: List[Tensor]) ``` There are a few problems exposed by HOP - AOTI expects a FX graph with weights as getattr nodes, aka stateful graph. HOP expect graph_module arguments to be stateless. Export serializer also expect a stateless graph. Currently, to make AOTI happy, I am making `original_gm` stateful, and bypassing the serialization for `original_gm`. - As a result, the HOP is not re-traceable, as functionalization on stateful graph module argument will fail. Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test Reviewed By: zhxchen17 Differential Revision: D68359391
3014224 to
cfb5471
Compare
67E6
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
Summary: Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering. I introduce a new HOP to address this. The schema is following ``` aoti_call_delegate(lower_moduel: AOTInductorEPModule, original_gm: fx.GraphModule, weights: List[Tensor], inputs: List[Tensor]) ``` There are a few problems exposed by HOP - AOTI expects a FX graph with weights as getattr nodes, aka stateful graph. HOP expect graph_module arguments to be stateless. Export serializer also expect a stateless graph. Currently, to make AOTI happy, I am making `original_gm` stateful, and bypassing the serialization for `original_gm`. - As a result, the HOP is not re-traceable, as functionalization on stateful graph module argument will fail. Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test Reviewed By: zhxchen17 Differential Revision: D68359391
cfb5471 to
133a4e8
Compare
|
This pull request was exported from Phabricator. Differential Revision: D68359391 |
|
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
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 |
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:
Previously, aoti compile node is represented as a kernel-less custom op in the exported program. The node was not eager runnable, which is a common practice for numerical validation during lowering.
I introduce a new HOP to address this.
The schema is following
There are a few problems exposed by HOP
original_gmstateful, and bypassing the serialization fororiginal_gm.Test Plan: buck2 test 'fbcode//mode/opt' fbcode//deeplearning/aot_inductor/cpu/test:cpu_lowering_utils_test
Reviewed By: zhxchen17
Differential Revision: D68359391