Add Multilingual TTS Support with ChatterboxMultilingualTTS

[4-alok]

🚀 Multilingual TTS Support Enhancement

This PR adds comprehensive multilingual text-to-speech capabilities to the Chatterbox TTS Server, enabling support for 23 languages including Hindi, Arabic, Chinese, French, German, and more.

✨ Key Features Added

🔄 Multilingual Model Integration

• Upgraded to ChatterboxMultilingualTTS from GitHub repository
• Supports 23 languages: Arabic (ar), Danish (da), German (de), Greek (el), English (en), Spanish (es), Finnish (fi), French (fr), Hebrew (he), Hindi (hi), Italian (it), Japanese (ja), Korean (ko), Malay (ms), Dutch (nl), Norwegian (no), Polish (pl), Portuguese (pt), Russian (ru), Swedish (sv), Swahili (sw), Turkish (tr), Chinese (zh)

🍎 Apple Silicon Compatibility

• Fixed MPS device loading issues for Apple Silicon Macs
• Resolved attention implementation conflicts with SDPA on MPS devices
• Added proper torch.load map_location handling for non-CUDA devices

🎛️ Enhanced User Interface

• Added model selection dropdown (Multilingual vs English-only)
• Integrated language selector with all 23 supported languages
• Added status badges showing current model type
• Improved UI state persistence and configuration management

🔧 Technical Improvements

• Updated server API endpoints (/tts, /v1/audio/speech) to handle language_id parameter
• Added graceful fallback for missing multilingual support
• Enhanced configuration system with multilingual settings
• Fixed various compatibility issues with newer PyTorch versions

📚 Documentation & Setup

• Added comprehensive documentation for multilingual setup
• Updated configuration examples and troubleshooting guides
• Included migration instructions for existing installations

🔧 Configuration Changes

The PR introduces new configuration options:

model:
  use_multilingual: true  # Enable multilingual model
  
generation_defaults:
  language: en  # Default language (English)

[hadessalvacion2025]

perfect.