ChatWithDB is an innovative project that transforms traditional databases into conversational partners using LLMs and Django. It accepts user input, finds similar database entries, and crafts human-like responses using OpenAI's GPT-3.5 or different LLMs, creating a unique interaction experience with databases. The project supports multiple databases, including PostgreSQL and MySQL.
- User-friendly Django admin dashboard to manage database connections.
- Integration with OpenAI's GPT-3 or Different LLMs to process user queries and responses.
- Advanced text similarity search with different databases.
- APIs to facilitate user interaction with databases.
First, clone the repository:
git clone https://github.com/shamspias/ChatWithDB.git
cd ChatWithDBInstall related pack
sudo apt-get install python3-dev python3-venv libcurl4-openssl-dev gcc libssl-dev -yCreate and active virtual environment
python3 -m venv venv
. venv/bin/activatepython -m venv venv
. venv/Scripts/activateInstall the required Python packages:
pip install -r requirements.txtConfigure your database in settings.py, then apply migrations:
python manage.py makemigrations
python manage.py migrateStart the Django Celery server:
celery -A config worker --loglevel=infoStart the Django development server:
python manage.py runserverThe server should be running on localhost:8000.
Navigate to the Django admin dashboard to add or manage your databases. Then, use the provided APIs to interact with your database. Enter your query, and the system will find similar entries in the database and craft a human-like response using GPT-3.5 or different LLMs.
We welcome contributions! Please see our Contributing Guidelines for more details.
This project is licensed under the terms of the MIT license. See the LICENSE file for details.