Add InternVL3 Guide

[congw729]

This PR adds the InternVL3 recipe as planned in #10

[gemini-code-assist]

Summary of Changes

Hello @congw729, 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 introduces a new guide for running the InternVL3 multimodal model series using vLLM. The guide provides comprehensive instructions for deployment, configuration, and validation, enabling users to efficiently set up and test InternVL3 on NVIDIA GPUs. This addition fulfills a planned recipe, expanding the available resources for model deployment within the repository.

Highlights

• New InternVL3 Usage Guide: This pull request introduces a new, detailed guide for deploying and utilizing the InternVL3 series of multimodal models on NVIDIA GPUs using vLLM.
• Comprehensive Deployment Steps: The guide provides clear, step-by-step instructions for installing vLLM and launching the InternVL3-8B-hf model, including specific commands for multi-GPU setups (e.g., A100-SXM4-40GB with tensor parallelism).
• Detailed Configuration Explanations: It explains crucial configuration parameters such as --enforce-eager (to disable CUDA Graph for compatibility), --tensor-parallel-size, and --data-parallel-size, helping users optimize their deployments.
• Validation and Benchmarking Instructions: The guide includes practical sections for validating the setup with a basic curl test and benchmarking performance using vllm bench serve, complete with expected output examples.

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 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 a usage guide for the InternVL3 model. The guide is well-structured and provides commands for setup, inference, and benchmarking. My review focuses on improving the clarity and correctness of the examples. I've suggested using a multimodal example to better showcase the model's capabilities and pointed out an inconsistency in the benchmark results. I've also made a few minor suggestions to improve the readability of the documentation.

[SamitHuang]

The Feature/benchmark/random mm data/images PR has been merged. I think we can replace the random data with multimodal random data based on it~

[congw729]

/gemini review

[gemini-code-assist]

Code Review

This pull request adds a new recipe for running InternVL3. The guide is well-structured and provides useful commands for setup, inference, and benchmarking. I've identified a couple of areas for improvement to enhance clarity and consistency. The main point is an inconsistency in the API endpoints used for basic testing versus benchmarking, which could be confusing. I've also suggested a minor text change to improve readability. Overall, this is a great addition.

[congw729]

@SamitHuang Already updated the test results for multimodal random dataset.

[congw729]

/gemini review

[gemini-code-assist]

Code Review

This PR adds a new recipe for InternVL3. The guide is comprehensive and covers installation, running the model, and benchmarking. I've found a few issues in the provided commands, mainly related to the benchmarking script, which will prevent them from running as-is. I've also suggested some minor formatting improvements for consistency and readability. Overall, a great addition once the corrections are made.

[SamitHuang]

I think it's better to set max concurrency to inf to test the max throughput

[SamitHuang]

can provide tips on how to set these arguments for random-mm

[congw729]

@ywang96 Hi, pls take a look. Thank u : )