… support

 MAJOR FEATURE: Complete Chatterbox TTS Integration

##  Core Implementation
-  Full Chatterbox TTS engine implementation in system/tts_engines/chatterbox/
-  Auto-model downloading via ChatterboxTTS.from_pretrained()
-  Cross-platform device support (CUDA/MPS/CPU) with automatic detection
-  Voice cloning via audio prompts from WAV files (no reference text required)
-  Low VRAM support with CPU/GPU memory swapping
-  Robust installation with dependency conflict resolution

##  Advanced Parameter System (5 New Controls)
-  **Seed**: Reproducible generation with cross-library seed setting
  - Supports torch, CUDA, random, and numpy seed synchronization
  - Smart random generation when seed=0 for varied outputs
-  **Temperature**: Neural network sampling randomness control
-  **Repetition Penalty**: Reduce repetitive speech patterns
-  **Exaggeration** (0.0-2.0): Speech expressiveness and emotion control
-  **CFG Weight** (0.0-1.0): Classifier-free guidance strength
-  **Min P** (0.0-1.0): Minimum probability threshold for sampling
-  **Top P** (0.0-1.0): Nucleus sampling parameter control

##  Complete API Integration
-  FastAPI endpoint updates in tts_server.py with new Form parameters
-  Dependency injection functions for default value handling
-  Parameter flow: Gradio  API  Model Engine  Chatterbox
-  Backward compatibility with existing AllTalk API structure
-  OpenAI API compatibility maintained

##  Enhanced Gradio Interface
-  5 new interactive sliders in Advanced TTS Settings
-  Smart capability-based visibility (only shown for Chatterbox)
-  Proper form submission and button handler updates
-  Real-time parameter validation and user feedback

##  Voice File System Integration
-  WAV file scanning in voices/ directory and subdirectories
-  Folder-based voice organization support
-  Audio prompt path resolution for voice cloning
-  Graceful fallback when voice files missing

##  Model Capabilities & Configuration
-  Updated model_settings.json with new parameter capabilities
-  Engine registration in tts_engines.json
-  Proper default values and validation ranges
-  Cross-platform compatibility flags

##  Technical Excellence
-  Comprehensive error handling with HTTP status codes
-  Debug logging with color-coded output
-  Memory management for CUDA environments
-  Generation locks to prevent concurrent conflicts
-  Proper async/await patterns throughout

##  Impact
- **Files Modified**: 7 core files across the AllTalk architecture
- **New Parameters**: 5 advanced generation controls
- **API Endpoints**: Updated with full parameter support
- **UI Components**: Enhanced with capability-aware controls
- **Compatibility**: Maintains full backward compatibility

This implementation transforms AllTalk into a more powerful TTS platform while maintaining its clean architecture and user-friendly interface. Chatterbox users now have professional-grade control over voice generation with industry-standard parameters.

… support

 MAJOR FEATURE: Complete Chatterbox TTS Integration

##  Core Implementation
-  Full Chatterbox TTS engine implementation in system/tts_engines/chatterbox/
-  Auto-model downloading via ChatterboxTTS.from_pretrained()
-  Cross-platform device support (CUDA/MPS/CPU) with automatic detection
-  Voice cloning via audio prompts from WAV files (no reference text required)
-  Low VRAM support with CPU/GPU memory swapping
-  Robust installation with dependency conflict resolution

##  Advanced Parameter System (5 New Controls)
-  **Seed**: Reproducible generation with cross-library seed setting
  - Supports torch, CUDA, random, and numpy seed synchronization
  - Smart random generation when seed=0 for varied outputs
-  **Temperature**: Neural network sampling randomness control
-  **Repetition Penalty**: Reduce repetitive speech patterns
-  **Exaggeration** (0.0-2.0): Speech expressiveness and emotion control
-  **CFG Weight** (0.0-1.0): Classifier-free guidance strength
-  **Min P** (0.0-1.0): Minimum probability threshold for sampling
-  **Top P** (0.0-1.0): Nucleus sampling parameter control

##  Complete API Integration
-  FastAPI endpoint updates in tts_server.py with new Form parameters
-  Dependency injection functions for default value handling
-  Parameter flow: Gradio  API  Model Engine  Chatterbox
-  Backward compatibility with existing AllTalk API structure
-  OpenAI API compatibility maintained

##  Enhanced Gradio Interface
-  5 new interactive sliders in Advanced TTS Settings
-  Smart capability-based visibility (only shown for Chatterbox)
-  Proper form submission and button handler updates
-  Real-time parameter validation and user feedback

##  Voice File System Integration
-  WAV file scanning in voices/ directory and subdirectories
-  Folder-based voice organization support
-  Audio prompt path resolution for voice cloning
-  Graceful fallback when voice files missing

##  Model Capabilities & Configuration
-  Updated model_settings.json with new parameter capabilities
-  Engine registration in tts_engines.json
-  Proper default values and validation ranges
-  Cross-platform compatibility flags

##  Technical Excellence
-  Comprehensive error handling with HTTP status codes
-  Debug logging with color-coded output
-  Memory management for CUDA environments
-  Generation locks to prevent concurrent conflicts
-  Proper async/await patterns throughout

##  Impact
- **Files Modified**: 7 core files across the AllTalk architecture
- **New Parameters**: 5 advanced generation controls
- **API Endpoints**: Updated with full parameter support
- **UI Components**: Enhanced with capability-aware controls
- **Compatibility**: Maintains full backward compatibility

This implementation transforms AllTalk into a more powerful TTS platform while maintaining its clean architecture and user-friendly interface. Chatterbox users now have professional-grade control over voice generation with industry-standard parameters.

… long inputs

- Added `chunk_text` method to split input text into manageable chunks at sentence boundaries, ensuring no overlaps.
- Updated TTS generation process to utilize text chunking, allowing for better handling of long texts (over 500 characters).
- Enhanced debug logging to verify chunking process and character counts for original and generated audio.
- Adjusted audio generation to concatenate chunks before saving, improving output quality and consistency.

…k_tts into alltalkbeta