Skip to content

Latest commit

 

History

History
19 lines (15 loc) · 986 Bytes

README.md

File metadata and controls

19 lines (15 loc) · 986 Bytes

Text2SQL-DjangoApp

A Django-based web application that converts natural language queries into SQL queries and executes them on user-uploaded datasets. This project integrates a fine-tuned T5 NLP model for accurate text-to-SQL conversion and supports REST API for seamless interaction between the frontend and backend.

Features

  • Upload Database: Users can upload their own database (in .db format).
  • Text-to-SQL Conversion: Converts natural language queries into SQL queries using a fine-tuned T5 Transformer model.
  • SQL Execution: Executes generated SQL queries on uploaded datasets and displays results.
  • REST API: Provides endpoints for dataset upload, query generation, and result fetching.
  • Model Training Notebook: Includes a Jupyter Notebook to fine-tune the T5 model on WikiSQL dataset, and evaluate its performance.

Requirements

The project requires the following dependencies:

  • Python 3.8+
  • Django 4.2+
  • PyTorch
  • Hugging Face Transformers
  • Pandas