| --- |
| 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. |
