Fix Torch ONNX export not respecting InputSpec.name #21646
Conversation
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @Flakes342, 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 addresses a critical usability issue in the Keras Torch backend's ONNX export functionality. Previously, exported models would have generic input names, making integration with downstream tools challenging. The changes ensure that user-defined input names from InputSpec are respected during the ONNX export, significantly improving the clarity and interoperability of exported models.
Highlights
- ONNX Export Input Naming Fix: Resolved an issue where the Torch backend's ONNX export process ignored
InputSpec.name, leading to generic input names likeargs_0in the exported model. Now,InputSpec.nameis properly propagated. - Fallback Naming Convention: Implemented a graceful fallback mechanism: if an
InputSpec.nameis not explicitly provided, the input will be namedinput_i(whereiis the input index), ensuring consistent and readable input names.
Using Gemini Code Assist
The 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 in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
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 .gemini/ folder in the base of the repository. Detailed instructions can be found here.
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
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
There was a problem hiding this comment.
Code Review
This pull request addresses an issue with ONNX export for the Torch backend where InputSpec.name was not being respected. The changes correctly extract the input names from the input_signature and pass them to torch.onnx.export. The fix is well-implemented and covers both the primary and fallback export paths. I have one suggestion to refactor the name extraction logic to be more concise and robust.
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #21646 +/- ##
=======================================
Coverage 82.55% 82.55%
=======================================
Files 571 571
Lines 57618 57646 +28
Branches 8999 9008 +9
=======================================
+ Hits 47564 47590 +26
- Misses 7759 7760 +1
- Partials 2295 2296 +1
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
Hi @fchollet, I think this PR is ready to be reviewed and merged now. It fixes the InputSpec functionality. I have formatted the code properly and also added a unit test for the same. |
Fixes #21637
Summary
Currently, when exporting a Keras model with the Torch backend to ONNX, the provided InputSpec.name is ignored. Instead, inputs are saved as generic names such as args_0, which makes exported models harder to use and integrate with downstream pipelines.
This PR updates the Torch backend export logic to properly propagate InputSpec.name into torch.onnx.export via the input_names argument. If a name is not provided, it gracefully falls back to the existing default (input_i).
Example
Example
Before:
ONNX model input name → args_0
After:
ONNX model input name → input
Please let me know what else is needed on this..