LitE-SQL: A Lightweight and Efficient Text-to-SQL Framework with Vector-based Schema Linking and Execution-Guided Self-Correction
Year: 2026
Type: conference-paper
Abstract: The Text-to-SQL task translates natural language questions into SQL queries, enabling intuitive database interaction for non-experts.While recent methods leveraging Large Language Models (LLMs) achieve strong performance, their reliance on proprietary models raises concerns about deployment feasibility and data privacy.In this work, we introduce LitE-SQL, a Lightweight and Efficient framework with two components: (i) a Schema Retriever that performs efficient schema linking using a vector database of pre-computed schema embeddings, optimized with a hardnegative supervised contrastive objective to distinguish semantically similar but functionally irrelevant columns, and (ii) a SQL Generator fine-tuned in two stages-supervised fine-tuning followed by execution-guided reinforcement-enabling execution-guided selfcorrection without multi-candidate sampling, which is commonly required by prior LLMbased approaches.On BIRD, LitE-SQL achieves 72.10% execution accuracy, and on Spider 1.0 it reaches 88.45%, demonstrating comparable or superior performance to LLMbased methods despite using 2× to 30× fewer parameters.Our findings demonstrate that high-quality Text-to-SQL generation is feasible with lightweight models, offering a practical solution for privacy-sensitive and resourceconstrained settings.Code is available at https://github.com/shengminp/LitE-SQL.
Authors: Shengmin Piao, Jieun Lee, Sanghyun Park
Institution: Yonsei University
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Awards: G2257911995, G4634960091
OpenAlex ID: W7140126946