fix ci and bug

[gushiqiao]

No description provided.

[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 primarily focuses on addressing several minor bugs and improving code robustness and clarity across different modules. Key changes include refining device assignment practices, enhancing the handling of LoRA weights and biases in convolutional layers, and optimizing model initialization logic for device placement. These adjustments aim to prevent potential runtime errors and streamline the codebase.

Highlights

• Device Assignment Refinement: Refactored device assignment logic in WeightModule to directly utilize AI_DEVICE, enhancing clarity and consistency by removing an intermediate target_device variable.
• Robust LoRA Weight/Bias Handling: Implemented robust handling for LoRA weight and bias application in Conv3dWeightTemplate by introducing dedicated getter methods (_get_actual_weight, _get_actual_bias), ensuring correct behavior even when diffs are not explicitly initialized.
• Optimized Model Initialization: Removed an unnecessary conditional CUDA device transfer during WanSFModel initialization, providing more flexible and controlled device management.
• Dependency Cleanup: Streamlined imports in intel_xpu.py by removing an unused torch.distributed module, reducing unnecessary dependencies.
• Minor Formatting Fixes: Applied minor formatting adjustments to configuration and Python files, including adding missing newlines for consistency.

🧠 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.

Changelog

• configs/z_image/z_image_turbo_t2i.json
  • Added a newline at the end of the file.
• lightx2v/common/modules/weight_module.py
  • Added a blank line after the AI_DEVICE import.
  • Removed the target_device local variable, directly using AI_DEVICE for parameter transfers to CUDA.
• lightx2v/common/ops/conv/conv3d.py
  • Removed the import of AI_DEVICE.
  • Removed the direct initialization of self.weight_diff and self.bias_diff to torch.tensor with AI_DEVICE.
  • Introduced _get_actual_weight and _get_actual_bias methods to safely retrieve the effective weight and bias, considering LoRA diffs.
  • Updated the apply method to utilize _get_actual_weight and _get_actual_bias for convolution operations.
• lightx2v/models/networks/wan/sf_model.py
  • Removed the conditional self.to_cuda() call from the __init__ method.
• lightx2v_platform/base/intel_xpu.py
  • Removed the torch.distributed import.
  • Added a newline at the end of the file.

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 introduces a series of fixes and refactorings across different components. The changes include fixing a bug in how LoRA-like diffs are applied in Conv3d, removing redundant variables and unused imports, and adjusting when models are moved to CUDA to improve flexibility. My review includes a suggestion to improve the readability of a newly introduced helper method in conv3d.py.

[gemini-code-assist]

The logic in _get_actual_bias can be simplified to improve readability and reduce duplication. The current implementation repeats the check for bias_diff. Consider refactoring to first determine the base bias and then apply the diff if it exists.

Suggested change

	def _get_actual_bias(self, bias=None):
	if bias is not None:
	if not hasattr(self, "bias_diff"):
	return bias
	return bias + self.bias_diff
	else:
	if not hasattr(self, "bias") or self.bias is None:
	return None
	if not hasattr(self, "bias_diff"):
	return self.bias
	return self.bias + self.bias_diff
	def _get_actual_bias(self, bias=None):
	if bias is None:
	if not hasattr(self, "bias") or self.bias is None:
	return None
	base_bias = self.bias
	else:
	base_bias = bias
	if hasattr(self, "bias_diff"):
	return base_bias + self.bias_diff
	return base_bias