Fix type/shape inferencing double recursion for subgraphs#29617
Open
skottmckay wants to merge 2 commits into
Open
Fix type/shape inferencing double recursion for subgraphs#29617skottmckay wants to merge 2 commits into
skottmckay wants to merge 2 commits into
Conversation
…on top of recursion for operators that do subgraph recursion via the ONNX type/shape inferencing path. For a model with deep nesting this results in model load taking an extreme amount of time. Memoize the subgraphs that have been processed to avoid this.
Contributor
There was a problem hiding this comment.
Pull request overview
This PR reduces pathological model-load time for deeply nested control-flow subgraphs by memoizing which subgraphs have already been type/shape-inferred (and therefore verified) during a Resolve pass, so the later “verify subgraphs” sweep can skip redundant VerifyNodeAndOpMatch recursion.
Changes:
- Track subgraphs that were processed via ONNX subgraph type/shape inference (e.g., Scan/If/Loop) in
ResolveContext. - Skip redundant
VerifyNodeAndOpMatchcalls for those already-processed subgraphs to avoid exponential re-traversal.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| onnxruntime/core/graph/graph.cc | Records subgraphs inferred via InferAndVerifySubgraphTypes and skips redundant subgraph verification during VerifyNodeAndOpMatch. |
| include/onnxruntime/core/graph/graph.h | Extends ResolveContext with an inferred_subgraphs set and clears it as part of context reset. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
The secondary recursion in VerifyNodeAndOpMatch results in recursion on top of recursion for operators that do subgraph recursion via the ONNX type/shape inferencing path.
For a model with deep nesting this results in model load taking an extreme amount of time.
Memoize the subgraphs that have been processed to avoid this.
Motivation and Context