fix bug

[chengtao-lv]

No description provided.

[gemini-code-assist]

Summary of Changes

Hello @chengtao-lv, 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 contains a targeted bug fix within the llmc/compression/token_reduction/divprune.py module. The change specifically resolves an indexing issue that could occur when determining the starting position of an image token, making the token reduction process more robust and preventing potential runtime errors in certain scenarios.

Highlights

• Bug Fix: I've addressed a bug in llmc/compression/token_reduction/divprune.py related to the extraction of the image_token_start_index. The previous implementation could fail if torch.where returned a tensor with multiple indices, as .item() expects a single-element tensor. The fix ensures that the first relevant index is reliably retrieved by adding an extra [0] before calling .item().

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 is currently in preview and 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 to provide feedback.

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

1. 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. ↩

[gemini-code-assist]

Code Review

The pull request fixes a bug when multiple image tokens are present. A suggestion has been made to also handle the case where no image token is found, which would currently cause an unhandled exception, to improve the code's robustness.

[gemini-code-assist]

If token_indices contains no True values, torch.where(token_indices)[0] will be an empty tensor. Accessing [0] on this empty tensor will raise an IndexError. Consider checking for this case and raising a more informative ValueError.

found_indices = torch.where(token_indices)[0]
if not found_indices.numel():
    raise ValueError("IMAGE_TOKEN_INDEX not found in input_ids, cannot perform DivPrune.")
pruning_paras['image_token_start_index'] = found_indices[0].item()