Audio-RAG Project

This project implements a Retrieval Augmented Generation (RAG) system for audio/video content. It allows you to transcribe multimedia files, index the resulting text into a vector database (ChromaDB), and query the database to find relevant information.

Setup

Follow these steps to set up the project on your system:

1. Clone the repository:

   git clone <repository URL>
   cd audio-rag # or the project directory name

2. Run the cross-platform setup script:

   python -m setup_project

   This script will create a virtual environment (.venv), install necessary dependencies, and guide you through configuring API keys in the .env file.

3. Activate the virtual environment:

   • On Linux/macOS:

     source .venv/bin/activate

   • On Windows:

     .venv\Scripts\activate

   Ensure the virtual environment is active whenever you work on the project (you will see (.venv) at the beginning of your terminal prompt).

Usage

Once the setup is complete and the virtual environment is activated, you can use the following scripts:

• setup_project.py: Performs the initial project setup (venv creation, dependency installation, .env configuration). Useful if you need to reconfigure the environment.

  python setup_project.py

• verify_setup.py: Verifies that the setup environment is correct (Python version, dependencies, FFmpeg, API keys, ChromaDB directory).

  python verify_setup.py

• app_query.py: Executes the logic for querying the indexed database.

  python app_query.py "Your query here"

• app_upload.py: Executes the logic for uploading and ingesting new audio/video files for transcription and indexing.

  python app_upload.py <path_to_audio/video_file>

• rag_system/verifier.py: Displays the content of the ChromaDB database.

  python rag_system/verifier.py

• rag_system/cleaner.py: Allows managing the ChromaDB database. Supports the following subcommands:

  • List documents:

    python rag_system/cleaner.py list

  • Delete specific documents by ID:

    python rag_system/cleaner.py delete <id1> <id2> ...

  • Delete the entire database:

    python rag_system/cleaner.py delete-all

Language Model Configuration

This project supports using either the Google Gemini model or a local model via Ollama.

To configure which model to use, set the LLM_TYPE environment variable in your .env file. It can be either gemini (default) or ollama.

Using Google Gemini

Set LLM_TYPE=gemini in your .env file and ensure your GOOGLE_API_KEY is set.

Using Ollama

1. Install Ollama from https://ollama.com/.
2. Pull the desired model (e.g., ollama pull llama2).
3. Set LLM_TYPE=ollama in your .env file.
4. Optionally, set OLLAMA_MODEL in your .env file to specify the model name (defaults to llama2).
5. Ensure you have installed the necessary dependencies by running python setup_project.py (this will install langchain-community).