Fetching metadata from the HF Docker repository...
-
Fix: Multilingual speech synthesis with advanced language detection and playback controls
Mark-Lasfar
-
Fix: Multilingual speech synthesis with advanced language detection and playback controls
Mark-Lasfar
-
Major update: Enhanced frontend and backend
Mark-Lasfar
- Added Mermaid diagram support for flowcharts and sequence diagrams
- Implemented Dark/Light theme toggle with localStorage persistence
- Fixed voice synthesis (speech) with voice loading detection
- Improved RTL/LTR support for Arabic/English
- Enhanced streaming performance in renderMarkdown
- Updated system prompt for deeper technical reasoning
- Switched to better models (DeepSeek-R1 / Qwen-72B)
- Improved mobile responsiveness and chat box sizing
- Added copy buttons for code blocks
- Fixed duplicate message prevention
- Optimized CSS for better performance
-
add chat
Mark-Lasfar
-
add chat
Mark-Lasfar
-
add chat
Mark-Lasfar
-
Fix flash-attn build by installing packaging early
