The problem AI SQL Query Builder solves

The AI SQL Query Builder solves the problem of simplifying the process of creating SQL queries for users without extensive SQL expertise. By interpreting natural language input and generating corresponding SQL queries, my tool makes retrieving specific information from a database easier. Additionally, integrating the tool with a database management system and showcasing the results in a web interface enhances user experience and facilitates the rapid development of database-driven applications. Furthermore, my SQLCoder 7B-2 model outperforms ChatGPT 4 in this task, highlighting the effectiveness and superiority of my approach in generating SQL queries from natural language input. Additionally, I have worked on another model, Llama 3 8B, to compare its accuracy with SQLCoder 7B-2 for the SQL Generator problem, which resulted in my decision to use two pre-trained models for the AI SQL Query Builder.

Challenges we ran into

The major problems we faced while creating this project was
1 getting accurate pre-trained models
for this, we had to check multiple models such as Gemma2B, Llama3 8B, nsql-Llama2 -7B, Sqlcoder 7B - 2
we checked the accuracy for all 4 and decided Sqlcoder was the best and after that it was llama3
2 Deploying the Project
we were confused about how to upload Such a large project Including the Models since the requirement was to use the models locally but we finally found out we could use docker and hugging face to upload our project
3 setup of the Gui
focusing on what we should use for creating the GUI and how to make it beautiful but finally, we decided on using Flask, Streamlit for the GUI respectively