- Notifications
You must be signed in to change notification settings - Fork3.5k
[quant] Fix QuantizeLinear opset-23 kernel registration to match ONNX schema#26491
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to ourterms of service andprivacy statement. We’ll occasionally send you account related emails.
Already on GitHub?Sign in to your account
Open
rachelElfenbein-dev wants to merge4 commits intomicrosoft:mainChoose a base branch
base:main
Could not load branches
Branch not found:{{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline, and old review comments may become outdated.
Open
[quant] Fix QuantizeLinear opset-23 kernel registration to match ONNX schema#26491
rachelElfenbein-dev wants to merge4 commits intomicrosoft:mainfromrachelElfenbein-dev:fix/quantizeLinear_Opset23
+25 −11
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
@microsoft-github-policy-service agree company="KamaTech" |
xadupre previously approved these changesNov 4, 2025
Sign up for freeto join this conversation on GitHub. Already have an account?Sign in to comment
Labels
None yet
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Uh oh!
There was an error while loading.Please reload this page.
Description
This PR fixes the type-constraint mapping for QuantizeLinear in opset 23 in the CPU EP kernel registration.
The kernel registration for opset 23 now matches the schema names and allowed types.
Type Constraints (opset 23)
Motivation and Context
Fixes#25932
Problem: Models with QuantizeLinear at opset 23 failed session initialization with “Could not find an implementation for QuantizeLinear(23)”, even though a CPU kernel exists.
The previous kernel registration for opset 23 incorrectly constrained the second input (y_scale, input index 1) to the quantized/output type (T).
As a result, at runtime KernelRegistry::TryFindKernel failed to match the kernel .
Root cause: The previous opset-23 registration did not mirror the schema’s per-parameter type variables. During KernelRegistry::TryFindKernel, the constraint-to-arg mapping (via KernelTypeStrResolver) compared Y_scale against the quantized type set (or otherwise mismatched sets), causing combinations to be rejected.
Solution: Register opset-23 with schema-accurate constraints (T1, T2, T3) so each argument is validated against the correct constraint . It restores the ability to find a matching kernel for valid models and removes the initialization failure.