Add streaming audio chunk support in Qwen3TTS model (#70)

* Add streaming audio chunk support in Qwen3TTS model

- Implemented a new method to decode audio chunks during speech generation.
- Updated `generateVoiceDesign` to yield audio chunks in a streaming manner.
- Added a helper function `decodeChunk` for processing codec codes into audio waveform.
- Enhanced audio generation flow to handle streaming efficiently, improving responsiveness and memory usage.

* Refactor TTSViewModel for improved streaming audio chunk handling

- Simplified streaming playback logic to allow audio chunks to be processed progressively.
- Removed unnecessary variable for audio data and streamlined chunk processing.
- Updated SettingsView to maintain UI consistency for streaming audio toggle based on chunking settings.