[Model] Add ToolParser and MoE Config for Hunyuan A13B #20820
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
👋 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 🚀 |
![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 @kzjeef, 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 integrating the Hunyuan A13B model with the system's tool calling and reasoning capabilities. It introduces a new tool parser tailored for Hunyuan A13B's specific output format and includes necessary adjustments to the core chat serving logic to support this new integration, particularly for streaming responses.
Highlights
- New Tool Parser: Introduced a dedicated
ToolParserfor the Hunyuan A13B model, enabling it to correctly parse and extract tool calls from the model's output, both in full and streaming modes. - Hunyuan Reasoning Integration: Ensured compatibility and proper functioning of the new tool parser with Hunyuan's existing reasoning parser, including minor fixes to improve its behavior.
- Streaming Output Enhancements: Improved the
serving_chat.pylogic to handle streaming tool call deltas more robustly, specifically addressing potentialNonevalue issues when concatenating token IDs and allowing tool parsers to modify the final message content.
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 introduces tool parsing support for the Hunyuan A13B model. My review focuses on improving the robustness and maintainability of the new parser. I've highlighted a potential high-severity issue with the regex for parsing nested JSON, and provided suggestions to make the code more concise and to refactor complex logic for better clarity.
vllm/entrypoints/openai/tool_parsers/hunyuan_a13b_tool_parser.py
Outdated
Show resolved
Hide resolved
vllm/entrypoints/openai/tool_parsers/hunyuan_a13b_tool_parser.py
Outdated
Show resolved
Hide resolved
|
This pull request has merge conflicts that must be resolved before it can be |
kzjeef
force-pushed
the
hy-tool-parser-submit
branch
from
a44378c to
d462300
Compare
mergify
bot
added
performance
kzjeef
force-pushed
the
hy-tool-parser-submit
branch
2 times, most recently
from
ede5da2 to
da46bfd
Compare
kzjeef
changed the title
kzjeef
requested review from
DarkLight1337,
robertgshaw2-redhat and
simon-mo
as code owners
|
I checked the entrypoints test, It meets error when startting a qwen2.5 - 1.5B model with length 8192, see log: and it's meets any error input too long for 8192. see: So how to change this test case 's length ? |
DarkLight1337
added
ready
- add stream and non stream support - reason parser use regex package. - reason parser: add missing function. Signed-off-by: Asher Zhang <asherszhang@tencent.com>
Signed-off-by: Asher Zhang <asherszhang@tencent.com>
- add test for hunyuan a13b tool parser. - fix mypy error on tool parser - refine reason parser test. - refactory tool parser stream function. Signed-off-by: Asher Zhang <asherszhang@tencent.com>
Signed-off-by: Asher Zhang <asherszhang@tencent.com>
- tune fused moe config. - benchmark: add hunyuan in moe benchmark Signed-off-by: Asher Zhang <asherszhang@tencent.com>
kzjeef
force-pushed
the
hy-tool-parser-submit
branch
from
da46bfd to
af5c48a
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.Purpose
Also fix some minor error in Hunyuan Reason parser.
Test Plan
Unit test:
OpenAI examples
Auto tool choice
openai client test without reason:
openai client test with reason
Test Result
Unit Test:
Note: nested json parameter in stream mode is not supported in this version, add a failure test case.
OpenAI examples
without reason, pass
tool_choicein following case:with reason, pass
tool_choicein following case:The require not work because they don't call tool parser, but the arguments have meta string like
<tool_calls></tool_calls>, which not filtered by chat server.(Optional) Documentation Update