Fix Wan/WanI2V patchification

[Jayce-Ping]

What does this PR do?

Fixes #13034

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline?
• Did you read our philosophy doc (important for complex PRs)?
• Was this discussed/approved via a GitHub issue or the forum? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

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.

[yiyixuxu]

thanks!

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[dg845]

@bot /style

[github-actions]

Style bot fixed some files and pushed the changes.

[dg845]

Thanks for the PR! Left a comment about using self.transformer.config.patch_size instead of hardcoding the patch_size to 2.

As an aside, I think this solution is superior to forcing the height and width to be multiples of 32 because WanPipeline is meant to work with both Wan 2.1 and Wan 2.2 models; because the VAE spatial downsample factor is 8 for the Wan 2.1 models but 16 for Wan 2.2 models, this check would be too restrictive for Wan 2.1 models (which can accept height/width which are multiples of 16 but not 32).

[Jayce-Ping]

@dg845 Hi! Thanks for your suggestion. Sorry that I hardcoded 2 instead of using patch_size. I agree with the idea that aut-resize is better than warn out.

Another potential issue (but unlikely happen) is that Wan2.2 has two transformers. What if in the future that some models have two transformers with different config.patch_size? But I think it is unlikely to happen since they need to have the same shape of latents to process. Anyway, the current solution is elegant enough.😊