Add qwen3-vl-dense support and example on Ascend

[HwVanICI]

Description

This PR adds support for training the Qwen3-VL dense models on Ascend NPU and provides training examples for Qwen3-VL-2B model on the Geometry3K dataset with GRPO.

Key change

Updated FSDP engine to correctly handle Qwen-VL model parameter names to ensure compatibility with vLLM-Ascend's internal model naming convention.

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not
  work as expected)
• Documentation update
• Code refactoring (no functional changes)
• Performance improvement
• Test coverage improvement

Checklist

• I have read the Contributing Guide
• I have run formatting tools (pre-commit or manual)
• I have run relevant unit tests and they pass
• I have added tests for new functionality
• I have updated documentation if needed
• My branch is up to date with main
• This PR introduces breaking changes (if yes, fill out details below)
• If this PR changes documentation, I have built and previewed it locally with
  jb build docs
• No critical issues raised by AI reviewers (/gemini review)

[gemini-code-assist]

Summary of Changes

Hello @HwVanICI, 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 expands the capabilities of the system by integrating support for Qwen3-VL dense models on Ascend NPUs. It includes crucial adjustments to the FSDP engine for proper model parameter handling and provides practical training examples, enabling users to leverage these advanced vision-language models effectively on the specified hardware.

Highlights

• Qwen3-VL Dense Model Support: This pull request introduces comprehensive support for training Qwen3-VL dense models on Ascend NPU, including new configuration files and examples.
• FSDP Engine Compatibility Fix: The FSDP engine has been updated to correctly handle Qwen-VL model parameter naming conventions, specifically addressing language_model and lm_head prefixes to ensure compatibility with vLLM-Ascend.
• Training Examples for Qwen3-VL-2B: New training examples are provided for the Qwen3-VL-2B model using the Geometry3K dataset with GRPO, making it easier to get started with this new model on Ascend.

🧠 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 adds support for Qwen3-VL dense models on Ascend, including updates to the FSDP engine for parameter name compatibility and new training examples. The changes look good overall. I've left a few minor comments regarding a typo in the documentation and some formatting suggestions in the new shell scripts to improve consistency.

[gemini-code-assist]

There appears to be a typo in the model name. To maintain consistency with the section title and configuration files, it should be Qwen2.5-VL-3B instead of Qwen2.5-VL-2B.

Suggested change

	1. **Qwen2.5-VL-3B** - `qwen2_5_vl_3b_geometry3k_grpo.*` - GRPO training on Geometry3K with Qwen2.5-VL-2B model dataset
	1. **Qwen2.5-VL-3B** - `qwen2_5_vl_3b_geometry3k_grpo.*` - GRPO training on Geometry3K with Qwen2.5-VL-3B model dataset

[garrett4wade]

Hi @HwVanICI , could you rebase the current main and resolve the conflict firest?

[garrett4wade]

LGTM

[HwVanICI]

Hi @HwVanICI , could you rebase the current main and resolve the conflict firest?

The branch is rebased to main.