fixed vlm ci error

[FrankLeeeee]

Motivation

Modifications

Related Issues

Accuracy Test

Benchmark & Profiling

Checklist

• Format your code according to the Code Formatting with Pre-Commit.
• Add unit tests as outlined in the Running Unit Tests.
• Update documentation / docstrings / example tutorials as needed, according to Writing Documentation.
• Provide throughput / latency benchmark results and accuracy evaluation results as needed, according to Benchmark and Profiling and Accuracy Results.
• For reviewers: If you haven't made any contributions to this PR and are only assisting with merging the main branch, please remove yourself as a co-author when merging the PR.
• Please feel free to join our Slack channel at https://sgl-fru7574.slack.com/archives/C09784E3EN6 to discuss your PR.

[gemini-code-assist]

Summary of Changes

Hello, 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 addresses a CI error related to VLM (Vision-Language Model) functionality and test execution. It includes a fix to ensure tensor device compatibility during VLM extension and adjusts test environment variable settings to prevent issues in the continuous integration pipeline, ultimately stabilizing the CI process.

Highlights

• VLM CI Fix: Modified the extend_vlm function in eagle3_target_model.py to explicitly move input_ids to CPU, resolving a potential device-related error in the VLM CI pipeline.
• Test Environment Configuration: Commented out the setting of os.environ variables for distributed training (RANK, WORLD_SIZE, MASTER_ADDR, MASTER_PORT) in test_dense within test_sglang_backend.py, likely to prevent conflicts or unnecessary configuration in the CI environment.

🧠 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 addresses a VLM CI error by ensuring a tensor is on the CPU before being passed to a function that expects it. While this change in specforge/modeling/target/eagle3_target_model.py appears correct, another change in tests/test_modeling/test_target/test_sglang_backend/test_sglang_backend.py comments out essential environment variable setup for a distributed test. This will likely break the test_dense function and seems to be an accidental change that should be reverted.

[gemini-code-assist]

The environment variable assignments for the distributed setup have been commented out. These are necessary for torch.distributed.init_process_group to function correctly, as it relies on the env:// initialization method by default. Other tests in this file, such as test_moe and test_vlm, retain this setup. This change will likely cause the test_dense test to fail. Please uncomment these lines.

Suggested change

	# os.environ["RANK"] = str(rank)
	# os.environ["LOCAL_RANK"] = str(rank)
	# os.environ["WORLD_SIZE"] = str(world_size)
	# os.environ["MASTER_ADDR"] = "localhost"
	# os.environ["MASTER_PORT"] = str(port)
	os.environ["RANK"] = str(rank)
	os.environ["LOCAL_RANK"] = str(rank)
	os.environ["WORLD_SIZE"] = str(world_size)
	os.environ["MASTER_ADDR"] = "localhost"
	os.environ["MASTER_PORT"] = str(port)