* refactor: update knowledge service and response structure
* feat: Enhance audio service with bi-streaming support and response models
- Introduced BiFlowableClientResponse interface to handle responses with separate body and stream item types.
- Updated ZaiClient to support bi-stream requests, allowing for more flexible streaming API interactions.
- Refactored AudioApi to rename methods for consistency and clarity, including audio transcription methods.
- Created new response classes: AudioCustomizationResponse, AudioSpeechResponse, and AudioTranscriptionResponse to encapsulate response data.
- Implemented AudioTranscriptionChunk and AudioTranscriptionResult classes for structured transcription results.
- Removed obsolete classes: AudioCustomizationApiResponse and AudioSpeechApiResponse.
- Updated AudioService and AudioServiceImpl to utilize new response models and handle streaming transcriptions.
- Enhanced unit tests for AudioService to reflect changes in method signatures and response types.
---------
Co-authored-by: tomsun28 <tomsun28@outlook.com>
* fix: update test cases
* feat: add new model constants and update AudioServiceTest to use them
* feat: update model list in README and README_CN with new AI capabilities
---------
Co-authored-by: tomsun28 <tomsun28@outlook.com>
