[Feature]: Update Emilia diarization model to pyannote/speaker-diarization-community-1

[gonzalo-cordova-pou]

Is your feature request related to a problem? Please describe.

preprocessors/Emilia/main.py currently hardcodes pyannote/speaker-diarization-3.1, while pyannote’s current open-source diarization pipeline is pyannote/speaker-diarization-community-1. This creates a mismatch with current pyannote recommendations and project setup docs.

Describe the solution you'd like

Update Emilia-Pipe to use pyannote/speaker-diarization-community-1 as the hardcoded diarization model in:

• preprocessors/Emilia/main.py

Also update matching docs in:

• preprocessors/Emilia/README.md setup instructions
• preprocessors/Emilia/README.md acknowledgement section

Describe alternatives you've considered

• Keep speaker-diarization-3.1 unchanged for strict historical reproducibility.
• Add a configurable model key in config.json.

For this contribution, the preferred approach is a minimal hardcoded replacement to match existing code style and keep the patch focused.

Additional context

In Pyannote words: Out of the box, Community-1 is much better than speaker-diarization-3.1.

References:

• https://huggingface.co/pyannote/speaker-diarization-community-1
• https://huggingface.co/pyannote/speaker-diarization-3.1
• https://github.com/pyannote/pyannote-audio
• https://docs.pyannote.ai/models
• https://www.pyannote.ai/blog/community-1