Feat: add support for multi speakers in AI logic voice config

[AustinBenoit]

Description

Provide details of the change, and generalize the change in the PR title above.

Firebase AI: Add support for advanced speech configurations, enabling multi-speaker setup and language code preferences in text-to-speech generation.

---

Testing

Describe how you've tested these changes.

Tested manually in play mode in unity.
Test with gemini and Vertex back ends for models
gemini-2.5-pro-preview-tts
gemini-2.5-flash-preview-tts
gemini-3.1-flash-tts-preview

Added a test for the serialization of the config data.

---

Type of Change

Place an x the applicable box:

• Bug fix. Add the issue # below if applicable.
• New feature. A non-breaking change which adds functionality.
• Other, such as a build process or documentation change.

---

[gemini-code-assist]

Code Review

This pull request introduces support for advanced speech configurations in Firebase AI, including multi-speaker setups and language code preferences in text-to-speech generation. It adds the necessary fields, structs, and serialization logic to GenerationConfig and LiveGenerationConfig, along with corresponding automated tests. The feedback suggests adding validation to prevent invalid serialization of default structs and ensuring that speaker and voice parameters are not null or empty to avoid bad API requests.

[milaGGL]

LGTM