|
--- |
|
title: SQL Generation |
|
emoji: 🦀 |
|
colorFrom: red |
|
colorTo: gray |
|
sdk: gradio |
|
sdk_version: 5.9.1 |
|
app_file: app.py |
|
pinned: false |
|
license: mit |
|
--- |
|
|
|
# SQL Generation 🦀 |
|
|
|
Welcome to the **SQL Generation** Gradio application! This tool leverages advanced machine learning models to assist in generating SQL queries based on natural language inputs. Whether you're a developer, data analyst, or just curious about SQL, this app aims to simplify the process of crafting SQL queries. |
|
|
|
## Features |
|
|
|
- **Natural Language to SQL**: Convert plain English descriptions into SQL queries. |
|
- **Multiple Datasets**: Trained on diverse datasets to handle various SQL generation tasks. |
|
- **User-Friendly Interface**: Built with Gradio for an intuitive and interactive experience. |
|
|
|
## Installation |
|
|
|
To run this application locally, ensure you have Python 3.10 or higher installed. Then, install the required dependencies: |
|
|
|
```bash |
|
pip install gradio transformers datasets |
|
``` |
|
## Usage |
|
After installing the dependencies, you can start the application by running: |
|
|
|
```bash |
|
python app.py |
|
``` |
|
This will launch a local server. Open your browser and navigate to http://127.0.0.1:7860 to access the interface. |
|
|
|
### Datasets Used |
|
The model has been trained on the following datasets: |
|
- b-mc2/sql-create-context: Provides context for SQL query generation. |
|
- TuneIt/o1-python: Offers examples of Python code snippets. |
|
- HuggingFaceFW/fineweb-2: Includes various language models for fine-tuning. |
|
- sentence-transformers/embedding-training-data: Supplies data for training sentence embeddings. |
|
|
|
## Model |
|
The application utilizes the distilbert-base-uncased model from Hugging Face, known for its efficiency and performance in natural language processing tasks. |
|
|
|
## License |
|
This project is licensed under the MIT License. |
|
|
|
## Acknowledgments |
|
- **Gradio** for providing an easy-to-use interface for machine learning models. |
|
- **Hugging Face** for hosting the pre-trained models and datasets. |
|
- **Datasets** for offering a wide range of datasets for training and evaluation. |
|
|
|
For more information, refer to the **Gradio** documentation. |
