solnone / GeminiSQLExecutor

This project is a Streamlit application that leverages Google’s Gemini model to translate English language questions into SQL queries to run on a local database.

GeminiSQLExecutor

Project Overview

This project is a Streamlit application that leverages Google’s Gemini model to translate English language questions into SQL queries to run on a local database.

Functions

The project consists of the following Python functions:

1. callGemini(question, prompt): This function uses the Gemini model to generate SQL queries from English language questions.
2. readSQL(query, db_user, db_password, db_host, db_name): This function executes the generated SQL query on a MySQL database and returns the results.
3. captureSchema(db_user, db_password, db_host, db_name): This function captures the schema of the MySQL database, which can be helpful for users to formulate their questions.

Flow of the Streamlit App

The Streamlit app is the front end of the project and interacts with the user:

1. Question Input: Users input their questions in English via a text area.
2. SQL Generation: Upon clicking a button, the callGemini function is invoked to generate an SQL query based on the user's question.
3. SQL Execution: The readSQL function is then invoked to execute the generated SQL query on a local MySQL database.
4. Result Display: The results of the SQL query execution are displayed in a table format.
5. Schema Inspection: The captureSchema function is used to provide a sidebar for inspecting the schema of the MySQL database.

This project takes advantage of:

• Natural Language Processing (NLP): The project uses Google's Gemini model, which is a state-of-the-art language model for generating SQL queries from English language questions.
• Web Development: The project uses Streamlit, a popular open-source app framework for Machine Learning and Data Science teams.
• Database Management: The project involves executing SQL queries on a MySQL database, providing hands-on experience with database management systems.
• Language Models: Google's Gemini model, a transformer-based language model, is used to generate SQL queries from English language questions.
• Multimodal Learning: Gemini's multimodal capabilities enable it to process and generate content from various sources, including text, images, and code.
• Transfer Learning: Gemini also benefits from transfer learning, where knowledge gained from pre-trained models on specific tasks is transferred to new tasks. This helps it learn faster and achieve better performance.