feat(diffusers): add lucy to diffusers #1390
Conversation
Summary of ChangesHello @Cui-yshoho, 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 Highlights
Using Gemini Code AssistThe 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
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 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
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces the LucyEditPipeline, a new video-to-video generation pipeline. The implementation is based on the existing WanPipeline but is adapted for video editing by concatenating conditional latents with the noise latents. The changes are well-structured, adding the new pipeline and its components.
My review focuses on improving performance and code clarity. I've identified an opportunity to significantly improve performance by batching the video encoding process. I've also pointed out some minor code cleanup opportunities, such as removing commented-out code and ensuring files end with a newline for consistency with coding standards.
| condition_latents = [ | ||
| retrieve_latents(self.vae, self.vae.encode(vid.unsqueeze(0))[0], sample_mode="argmax") for vid in video | ||
| ] | ||
|
|
||
| condition_latents = mint.cat(condition_latents, dim=0).to(dtype) |
There was a problem hiding this comment.
The current implementation encodes each video in the batch individually using a list comprehension. This is inefficient as it prevents batch processing and can be very slow, especially for larger batches. The VAE encoder is capable of processing a whole batch of videos at once, which is much more performant.
condition_latents = retrieve_latents(self.vae, self.vae.encode(video)[0], sample_mode="argmax").to(dtype)|
|
||
| # latent_model_input = latents.to(transformer_dtype) | ||
| latent_model_input = mint.cat([latents, condition_latents], dim=1).to(transformer_dtype) | ||
| # latent_model_input = mint.cat([latents, latents], dim=1).to(transformer_dtype) |
There was a problem hiding this comment.
This commented-out line appears to be a leftover from debugging or experimentation. It should be removed to improve code clarity and maintainability.
| if not return_dict: | ||
| return (video,) | ||
|
|
||
| return LucyPipelineOutput(frames=video) No newline at end of file |
There was a problem hiding this comment.
The file is missing a final newline character. It's a good practice to end files with a newline for POSIX compatibility and to avoid issues with some tools.
return LucyPipelineOutput(frames=video)| ) | ||
|
|
||
| for name, value in _dummy_objects.items(): | ||
| setattr(sys.modules[__name__], name, value) No newline at end of file |
There was a problem hiding this comment.
The file is missing a final newline character. It's a good practice to end files with a newline for POSIX compatibility and to avoid issues with some tools.
| setattr(sys.modules[__name__], name, value) | |
| setattr(sys.modules[__name__], name, value) | |
| `(batch_size, num_frames, channels, height, width)`. | ||
| """ | ||
|
|
||
| frames: ms.Tensor No newline at end of file |
There was a problem hiding this comment.
The file is missing a final newline character. It's a good practice to end files with a newline for POSIX compatibility and to avoid issues with some tools.
| frames: ms.Tensor | |
| frames: ms.Tensor | |
1 participant
What does this PR do?
Fixes # (issue)
Adds # (feature)
Before submitting
What's New. Here are thedocumentation guidelines
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@xxx