AudioNotes is a system for transcribing and summarizing audio/video content into structured Markdown notes, leveraging FunASR for speech-to-text and Qwen2 for large language model-based organization. It targets users who need to quickly process and digest spoken or recorded information, providing an efficient way to create readable summaries and interact with the content.

How It Works

The system integrates FunASR for accurate speech recognition and Qwen2 (specifically the 7B parameter model) for advanced natural language processing. Audio or video files are first transcribed by FunASR. The resulting text is then processed by Qwen2, which organizes the information into a structured Markdown format, enabling users to quickly review and understand the content.

Quick Start & Requirements

• Installation:
  • Docker (Recommended):

    curl -fsSL https://github.com/harry0703/AudioNotes/raw/main/docker-compose.yml -o docker-compose.yml
    docker-compose up
    

    Access at http://localhost:15433/ (default credentials: admin/admin).
  • Local:

    conda create -n AudioNotes python=3.10 -y
    conda activate AudioNotes
    git clone https://github.com/harry0703/AudioNotes.git
    cd AudioNotes
    pip install -r requirements.txt
    # Rename .env.example to .env and configure
    chainlit run main.py
    

    Access at http://localhost:8000/ (default credentials: admin/admin).
• Prerequisites: Ollama installed and Qwen2 model pulled (ollama pull qwen2:7b). Local deployment requires a PostgreSQL database.
• Resources: Requires Ollama and a Qwen2 model. Docker deployment simplifies dependency management.

Highlighted Details

• Leverages FunASR for speech-to-text and Qwen2 for content summarization.
• Outputs structured Markdown notes for easy readability.
• Offers both Docker and local deployment options.
• Supports interaction with audio/video content through Qwen2.

Maintenance & Community

The project is maintained by harry0703. Further community or roadmap information is not detailed in the README.

Licensing & Compatibility

The README does not explicitly state the license. Compatibility for commercial or closed-source use is not specified.

Limitations & Caveats

The project appears to be actively developed, with potential for breaking changes. Specific performance benchmarks or detailed limitations are not provided in the README.