Adding support for custom ort tokenizer ops and ai.onnx.ml normalizer#34721
Merged
sfatimar merged 26 commits intoMay 1, 2026
Merged
Conversation
Contributor
Author
|
@mvafin please review |
Contributor
|
Please, fix python formatting. |
Contributor
There was a problem hiding this comment.
Pull request overview
This PR aims to extend the ONNX frontend with (1) ai.onnx.ml Normalizer operator support and (2) improved guidance when encountering unsupported tokenizer-related ops by pointing users to OpenVINO Tokenizers extensions. It also adds new ONNX layer tests intended to cover these ops.
Changes:
- Add ai.onnx.ml
Normalizertranslator implementation and register it under theai.onnx.mldomain. - Add an additional error callback to surface “install openvino-tokenizers” guidance when specific tokenizer ops remain unconverted.
- Add new ONNX layer tests for Tokenizer / TfIdfVectorizer / StringNormalizer / LabelEncoder.
Reviewed changes
Copilot reviewed 9 out of 9 changed files in this pull request and generated 5 comments.
Show a summary per file
| File | Description |
|---|---|
| tests/layer_tests/onnx_tests/test_tokenizer.py | Adds a new nightly test for com.microsoft::Tokenizer (currently committed with invalid Python syntax). |
| tests/layer_tests/onnx_tests/test_tfidf_vectorizer.py | Adds a new nightly test for TfIdfVectorizer (currently committed with invalid Python syntax). |
| tests/layer_tests/onnx_tests/test_string_normalizer.py | Adds a new nightly test for StringNormalizer (currently committed with invalid Python syntax). |
| tests/layer_tests/onnx_tests/test_label_encoder.py | Adds a new nightly test for ai.onnx.ml::LabelEncoder (currently committed with invalid Python syntax). |
| src/frontends/onnx/frontend/src/ops_bridge.hpp | Declares helper to list tokenizer ops for error reporting. |
| src/frontends/onnx/frontend/src/ops_bridge.cpp | Defines AIONNX_ML_DOMAIN and implements get_supported_ops_via_tokenizers(). |
| src/frontends/onnx/frontend/src/frontend.cpp | Adds tokenizer-aware check_unconverted_ops additional callback messaging. |
| src/frontends/onnx/frontend/src/core/operator_set.hpp | Exposes AIONNX_ML_DOMAIN declaration. |
| src/frontends/onnx/frontend/src/op/ai.onnx.ml/normalizer.cpp | Adds translator for ai.onnx.ml::Normalizer (currently has correctness issues vs ONNX semantics). |
Contributor
Author
|
@mvafin can we please merge this PR |
Contributor
Author
|
@mvafin I have made clang format changes. Can you please restart the pipelines. |
Contributor
Author
|
@mvafin please start pipelines again |
Contributor
|
@mvafin Please trigger ci/jenkins pipeline |
Contributor
|
build_jenkins |
mvafin
approved these changes
Apr 21, 2026
Contributor
Author
|
@openvinotoolkit/openvino-tests-maintainers Can you please review this PR |
mvafin
requested changes
Apr 28, 2026
mvafin
left a comment
mvafin
left a comment
Contributor
There was a problem hiding this comment.
Please remove tests from tests/layer_tests/onnx_tests. They don't run and we cannot be sure if they are passing or not.
mvafin
approved these changes
Apr 28, 2026
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.
Details:
Tickets:
(https://jira.devtools.intel.com/browse/CVS-181364)
AI Assistance: