[Doc] Update V1 status for decoder-only embedding models #19952
Conversation
Signed-off-by: Isotr0py <2037008807@qq.com>
Signed-off-by: Isotr0py <2037008807@qq.com>
|
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
There was a problem hiding this comment.
Summary of Changes
Hello @Isotr0py, 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 primarily focuses on enhancing the project's documentation by updating the V1 compatibility status for various model architectures in the supported_models.md file. Concurrently, it includes a minor but important code cleanup within the Qwen2 Process Reward Model implementation to improve its structure and alignment with standard practices.
Highlights
- Documentation Update: I've updated the
docs/models/supported_models.mdfile to reflect the V1 compatibility status for several model architectures across embedding, reward, classification, and scoring tasks. This includes marking Gemma2, Llama, Qwen2, Qwen3 (embedding), InternLM2, Llama, Qwen2 (reward), GPT2 (classification), and Qwen3 (scoring) as V1 compatible. - Code Refinement: I've performed a cleanup of the
Qwen2RewardBaseModelimplementation invllm/model_executor/models/qwen2_rm.py. This involved removing theSupportsV0Onlyinterface, replacing a customReLUclass with PyTorch's built-innn.ReLU, and explicitly settingreturn_bias=Falsefor the linear layers within the scoring head.
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
-
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 updates the V1 support status for several models in the documentation and refactors the Qwen2 reward model implementation. The documentation changes are accurate and align with the PR's goal. However, the refactoring in vllm/model_executor/models/qwen2_rm.py introduces a critical issue that will cause a runtime error. I've provided a comment with a suggested fix to address this.
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: juncheoll <th6re8e@naver.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: fhl <2410591650@qq.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: Will Eaton <weaton@redhat.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: avigny <47987522+avigny@users.noreply.github.com>
…t#19952) Signed-off-by: Isotr0py <2037008807@qq.com>
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.Purpose
Test Plan
Test Result
(Optional) Documentation Update
docs/models/supported_models.md