Abstract
The transformation of natural language text into SQL queries is a critical task in the domain of natural language processing and database management. This paper presents a comparative analysis of three prominent models: T5 (Text-to-Text Transfer Transformer), Seq2Seq (Sequence-to-Sequence), and SQLNet, applied to the Spider and WikiSQL dataset—a complex and diverse benchmark for text-to-SQL tasks. Leveraging the power of Large Language Models (LLM), we explore the efficacy, accuracy, and generalization capabilities of these models in transforming natural language queries into executable SQL statements