feat: add support for Precision-2 (PYA-964)

[hbredin]

No description provided.

[Copilot]

Pull Request Overview

This PR adds support for Precision-2 model by introducing the new model as the default for diarization, voiceprint, and identification operations. The changes also enable previously unsupported speaker count constraints and add new confidence and exclusivity options.

• Updates type annotations from Optional[int] to int | None for modern Python syntax
• Adds support for min_speakers and max_speakers parameters that were previously disabled
• Introduces new parameters: turn_level_confidence, exclusive, and model with "precision-2" as default

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The documentation format is inconsistent. The model parameter should include a period at the end to match the format of other parameter descriptions in the docstring.

Suggested change

	Defaults to "precision-2"
	Defaults to "precision-2".

[Copilot]

The model parameter documentation is incomplete - it's missing a description of what the model parameter does, unlike other parameters that have descriptive text explaining their purpose.

Suggested change

	Defaults to "precision-2".
	Name of the model to use for speaker identification. Defaults to "precision-2".

[Copilot]

The new parameters turn_level_confidence and exclusive lack descriptions explaining their purpose and functionality, making the API unclear to users.

Suggested change

	Defaults to False
	exclusive: bool, optional
	Defaults to False
	If True, includes confidence scores for each detected speech turn in the diarization output.
	Defaults to False.
	exclusive: bool, optional
	If True, ensures that speech turns do not overlap between speakers, enforcing exclusivity in speaker assignment.
	Defaults to False.