Text-to-SQL Generator (TinyLLaMA + LoRA + Flask)

This project is a Flask-based web application that converts natural language questions into SQL queries using a LoRA-fine-tuned TinyLLaMA model. It utilizes the Synthetic-Text-To-SQL dataset for finetuning and is designed to run locally or via a public URL using ngrok.

---

Features

• Converts natural language questions to executable SQL queries
• Fine-tuned with LoRA (Low-Rank Adaptation) for efficient adaptation
• Uses TinyLLaMA for lightweight inference
• Easy deployment with Flask + pyngrok
• Clean, responsive HTML UI

---

Getting Started

1. Clone the Repository

git clone https://github.com/SwathiR1999/Text2Sql.git
cd Text2Sql

2. Install requirements

pip install -r requirements.txt

3. Configure pyngrok

Visit ngrok to get your auth token.

from pyngrok import ngrok
ngrok.set_auth_token("YOUR_NGROK_AUTH_TOKEN")

3. Run the app

python app.py

You will see

* ngrok tunnel: <...link...>

Open the link in your browser to access the web app.