In Chinese culture, people will write songs or poetry for someone they love. So we want to use AI to record life stories and convey love through a music album. You only need to share the backstories or experiences of the character, and then customised music album will be produced album image, Album songs lyrics, enjoying an audio-visual feast.
At present, we have only done a local deployment version. Because some APIs need to be paid, they are not open to the public for the time being. I will update it later when it is launched online.
Features
- Select character stories from YAML-formatted character library
- Generate professional album covers based on character information
- Create personalized lyrics and stories by analyzing character information
- Recommend the most suitable music style
- Generate high-quality music works
- Play and manage generated albums through a modern web interface
- Display music trivia while waiting for generation
- Support real-time progress display and status updates
- Provide album history and recent generation list
Note: Character information upload feature is under development and will be available in future versions. Currently only supports selecting characters from the preset character library.
Technology Stack
- Frontend: HTML5, CSS3, JavaScript, jQuery, Bootstrap 5
- Backend: Python Flask 3.0.2
- AI Models:
- Kolors (Album cover generation)
- Qwen2.5-32B-Instruct (Lyrics and style generation, plus music trivia)
- Suno Chirp-v3-5 (Music generation)
- Data Storage: YAML files, Local file system
- Dependency Management: pip + requirements.txt
System Requirements
- Python 3.8+
- Operating System: Windows/macOS/Linux
- Memory: Minimum 4GB RAM
- Disk Space: Minimum 2GB available space
- Network: Stable internet connection (for API calls)