sqlrag

sqlrag helps users to query SQL databases using natural language. It leverages LLM and embeddings to interpret user queries and generate appropriate SQL statements.

Features

• Natural language interface for SQL queries
• Returns both query results and the generated SQL statement
• Currently only supports SQLite database
• Currently only supports Ollama for language model and embeddings

Note: This project is currently under development.

Setup and Running

1. Clone the repository:

   git clone https://github.com/yourusername/sqlrag.git
   cd sqlrag
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Configure .env file:

   DB = <sqlite_db_name>
   LLM_MODEL = <ollama_model_name>
   EMBEDDING_MODEL = <ollama_embedding_model_name>
   

4. Run the script:

   python sqlrag.py
   

Install Ollama

You can follow the instructions on the Ollama GitHub repository to install and run Ollama on your system.

Sample SQLite database:

Create:

python create_database.py

Check:

python check_tables.py

TODO

• Add support for SQLite database
• Add support for CSV file
• Add support for PostgreSQL database
• Add support for MySQL database
• Implement basic web interface
• Integrate basic data visualization
• Add query history tracking
• Add basic authentication
• Data validation for CSV files
• Multiple CSV files support

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.