[Qwen] avoid creating attention masks when there is no padding

[kashif]

What does this PR do?

We now have a unified Qwen prompt-embed handling to avoid creating attention masks when there is no padding (keeps flash SDPA eligible, removes unnecessary masks). This fixes the inference speed regression.

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.

[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.

[sayakpaul]

Thanks!

Could you update with the following things?

• Shed light into what caused the speed regression
• Add a test with masks in the compilation tests here
• Do a before and after comparison in the outputs with the PR

[kashif]

will do!

[yiyixuxu]

curiously, cant we just do

if encoder_attention_mask.all():
    encoder_attention_mask = None

[kashif]

the issue was that this was giving trouble with torch.compile...

[yiyixuxu]

ohh, but I think we typically just torch.compile the models (e.g. pipeline.transformer.compile(...), pipeline.text_encoder.compile())
the computations on the pipeline won't be part of it though

[kashif]

oh ok let me check that then