Correctness Check SQL Validation with BERT

The model must be trained with our predefined solution for each question.
● It should be capable of comparing user inputs with the solution, checking for both content and syntax errors.
● The integration of various schema comparison methods, especially the CUPID method, is crucial for providing precise feedback.

The aim is to develop a robust and reliable model that helps users improve their SQL skills through practical exercises. The model should not only differentiate between correct and incorrect SQL but also provide constructive feedback to enable users to learn from their mistakes