[Relax] Add FDataDependent operator attribute for LegalizeOps #18664
Conversation
Summary of ChangesHello @guan404ming, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request refactors the identification of data-dependent operators within the Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request is a great improvement, replacing fragile string matching for data-dependent operators with a dedicated FDataDependent attribute. This makes the LegalizeOps transform more robust and scalable. The changes in src/relax/op/tensor/index.cc to add the attribute and in tests/python/relax/test_transform_legalize_ops_index_linear_algebra.py to test it are clear and correct. I have one suggestion in src/relax/transform/legalize_ops.cc to simplify the attribute checking logic for better readability and conciseness.
| bool is_data_dependent_op = [&]() -> bool { | ||
| if (Op::HasAttrMap("FDataDependent")) { | ||
| auto op_map = Op::GetAttrMap<Bool>("FDataDependent"); | ||
| if (op_map.count(op)) { | ||
| return op_map[op]->value; | ||
| } | ||
| } | ||
| return false; | ||
| }(); |
There was a problem hiding this comment.
This immediately-invoked lambda expression can be simplified to a more concise and idiomatic one-liner using Op::GetAttrMap and the .get() method with a default value. This will improve readability.
bool is_data_dependent_op = Op::GetAttrMap<Bool>("FDataDependent").get(op, Bool(false))->value;
1 participant
Why
The LegalizeOps transform was using string matching to detect data-dependent operators by checking if "dynamic" appears in the operator name. This approach is fragile and doesn't scale well as new data-dependent operators are added.
How