[fix] Make Qwen2_5OmniProcessor warning a lot less noisy via warning_once

[tomaarsen]

What does this PR do?

There's 2 changes, one is a definite fix and one is a preference. Some background: there are a lot of models that have finetuned qwen2_5_omni, e.g. https://huggingface.co/LCO-Embedding/LCO-Embedding-Omni-3B, and in the case of this model, finetuned with a whole new chat template/system prompt/no system prompt.

This is totally fine, except every single input processing results in a spam of

WARNING:root:System prompt modified, audio output may not work as expected. Audio output mode only works when using default system prompt 'You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.'
WARNING:root:System prompt modified, audio output may not work as expected. Audio output mode only works when using default system prompt 'You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.'
WARNING:root:System prompt modified, audio output may not work as expected. Audio output mode only works when using default system prompt 'You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.'
WARNING:root:System prompt modified, audio output may not work as expected. Audio output mode only works when using default system prompt 'You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.'
...

So I've made 2 changes:

1. Use warning_once instead of warning every single time
2. (My preference) being able to disable the warning altogether by setting check_audio_system_prompt or check_system_prompt in the processor config. That way I should be able to remove the warning altogether in these cases where it's a totally incorrect warning.

If change 2 goes too far, I can revert that commit and we can go for just option 1.

An alternative solution would be to just remove the warning: we don't often get involved with the chat template/system prompt choices of the architectures, right?

Code Agent Policy

• I confirm that this is not a pure code agent PR.

This one was written manually, except the test, which was half generated

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,
  Pull Request section?
• 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?

@zucchini-nlp

• Tom Aarsen

[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.

[zucchini-nlp]

Yes, let's delete entirely!

[zucchini-nlp]

Oh I hate this warning as well haha, couldn't convince qwen team to delete at model addition time 🫠

Now that it's maintained by us, ig we can decide what to do with it

[tomaarsen]

That's music to my ears! Removed it fully now.

They don't make PRs better than this:

• Tom Aarsen

[zucchini-nlp]

Yaay, we can delete apply_chat_template entirely and merge!

[zucchini-nlp]

we can completely delete the whole fn 💃🏻

[tomaarsen]

😲

[tomaarsen]

Done in e0d7086

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: qwen2_5_omni, qwen3_omni_moe