README.md

metadata

title: NL2SQL API
emoji: 🗄️
colorFrom: blue
colorTo: indigo
sdk: docker
pinned: false

NL2SQL

A backend Command Line Interface (CLI) framework designed to evaluate and test various NL2SQL models.

Note: The frontend interface for this application is currently in progress and not yet available. All interactions are handled via the CLI.

Prerequisites & Installation

To run this CLI tool locally, follow these steps to set up your environment:

1. Activate the Virtual Environment.

• Windows: venv\Scripts\Activate
• macOS/Linux: source venv/bin/activate

2. Install Requirements

• Ensure a requirements.txt file exists in your project backend folder.

  pip install -r requirements.txt

3. Configure Environment Variable (.env)

Note: Users must generate your own free access token from your https://huggingface.co/settings/tokens to avoid rate limits.

• Create a .env file in the backend/ directory

• Add your Hugging Face Read Token to the file.

  HF_TOKEN=your_hugging_face_read_token_here

Usage Guide

Once your environment is set up and your token is configured, you can run the CLI application.

1. Navigate to the Backend Directory

   cd backend

2. Launch the Application

   python -m app.main

3. Interacting with the CLI Menu Upon running the command, you will be presented with a main menu. Choose the number corresponding to your desired action:

   1. Running Question to SQL Test: Evaluates how well a model translates natural language queries into executable SQL commands.

   2. Running Question Answering Test: Evaluates the end-to-end process (Question -> SQL -> Database Execution -> Natural Language Answer).

   3. Exit/Quit: Closes the application.

4. Model Selection & Batch Testing

   1. After selecting either option 1 or 2, the CLI will display a list of available NL2SQL models.

   2. Enter the number/name of the model you wish to test.

   3. Automatic Execution: Once a model is selected, the system will automatically begin running the batch test against the scenarios defined in scripts/test_cases.json. Sit back and wait for the reports to generate in your root folder!

🚧 Roadmap

• Development and integration of a graphical User Interface (Frontend).

• Additional database schema support.