[bugfix] fix lora in vllm >= v0.12

[liuyanyi]

PR type

• Bug Fix
• New Feature
• Document Updates
• More Models or Datasets Support

PR information

Since vllm has remove extra vocab for LoRA, lora_extra_vocab_size access led to error in training, use get_attr to support newer version. Tested on vllm v0.13.0

Experiment results

Paste your experiment result here(if needed).

[gemini-code-assist]

Summary of Changes

Hello @liuyanyi, 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 resolves a bug that prevented LoRA (Low-Rank Adaptation) functionality from working correctly with vLLM versions 0.12 and above. The changes adapt the LoRA adapter loading mechanism to account for the removal of the lora_extra_vocab_size attribute in newer vLLM releases, ensuring seamless integration and preventing errors during model training.

Highlights

• VLLM Compatibility: The LoRA loading logic has been updated to maintain compatibility with vLLM versions 0.12 and newer, which have removed the lora_extra_vocab_size attribute.
• Dynamic Parameter Handling: Conditional checks using hasattr have been introduced for embedding_padding_modules and lora_extra_vocab_size to ensure that the correct parameters are passed to LoRA loading functions, adapting to different vLLM versions.
• Code Refactoring: Common arguments for from_lora_tensors and from_local_checkpoint methods are now consolidated into a lora_request_kwargs dictionary, improving code readability and maintainability.

🧠 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 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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request 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

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

This pull request correctly addresses a compatibility issue with lora_extra_vocab_size in newer vLLM versions by conditionally accessing the attribute. The refactoring of LoRA request arguments into a dictionary is a good improvement for code clarity and maintainability. I have one suggestion to improve the exception handling pattern.

[gemini-code-assist]

Using raise e can obscure the original traceback, making debugging more difficult. It's better to use a bare raise to re-raise the caught exception while preserving its full traceback.

Suggested change

	except Exception as e:
	raise e
	except Exception:
	raise

[hjh0119]

LGTM
thanks!