Add function calling support to DeepSeek-V3.1 usage guide.

[Evrard-Nil]

This pull request updates the documentation for running DeepSeek-V3.1 models with vLLM, specifically adding instructions for enabling function calling support. The main change is a new section describing how to configure vLLM to support user-defined functions.

Function calling support:

• Added a "Function calling" section to DeepSeek/DeepSeek-V3_1.md with instructions for enabling function calling in vLLM using DeepSeek-V3.1 models, including the necessary command-line arguments (--enable-auto-tool-choice, --tool-call-parser deepseek_v31, and --chat-template /models/DeepSeek-V3.1/assets/chat_template.jinja).

This is based on this PR.

[gemini-code-assist]

Summary of Changes

Hello @Evrard-Nil, 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 significantly improves the documentation for DeepSeek-V3.1 models by integrating instructions for enabling function calling support within vLLM. It provides clear guidance on the necessary configuration parameters, empowering users to leverage advanced tool-use capabilities with these models.

Highlights

• Documentation Update: A new section has been added to the DeepSeek/DeepSeek-V3_1.md documentation.
• Function Calling Support: The new section provides detailed instructions on how to enable function calling for DeepSeek-V3.1 models when running with vLLM.
• vLLM Configuration: Specific command-line arguments (--enable-auto-tool-choice, --tool-call-parser deepseek_v31, --chat-template) are outlined for configuring vLLM to support function calling.

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

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 adds documentation for function calling with DeepSeek-V3.1. The changes are a good addition, but I have a couple of suggestions to improve clarity for users. The example command could be more specific to avoid ambiguity, and the hardcoded path for the chat template could be confusing. I've added comments with specific suggestions to address these points.

[gemini-code-assist]

The vllm serve ... syntax is a bit ambiguous, and using ... on a separate line is unconventional. It would be clearer to show a more complete command that includes the model name, to indicate that these are additional arguments, similar to the previous example in the document.

Suggested change

	vllm serve ...
	vllm serve deepseek-ai/DeepSeek-V3.1 \

[gemini-code-assist]

The hardcoded path /models/DeepSeek-V3.1/assets/chat_template.jinja may be confusing for users, as it assumes a specific directory structure that might not exist on their system. It would be more helpful to use a placeholder and explain where to find the chat_template.jinja file.

You could add a note like this to the documentation before the code block:
"Note: The chat_template.jinja file is required for function calling. You can find it in the vLLM source code at vllm/assets/deepseek_v31/chat_template.jinja. You need to provide the correct path to this file."

Suggested change

	--chat-template /models/DeepSeek-V3.1/assets/chat_template.jinja
	--chat-template <path/to/your/chat_template.jinja>