MSc-SQL: Multi-Sample Critiquing Small Language Models For Text-To-SQL Translation

Satya Krishna Gorti, Ilan Gofman, Zhaoyan Liu, Jiapeng Wu, Noël Vouitsis +3 more
10/16/2024
cs.CL

Official code: layer6ai-labs/msc-sql (19 stars, MIT, last commit 2025-02-21).

Abstract

Text-to-SQL generation enables non-experts to interact with databases via natural language. Recent advances rely on large closed-source models like GPT-4 that present challenges in accessibility, privacy, and latency. To address these issues, we focus on developing small, efficient, and open-source text-to-SQL models. We demonstrate the benefits of sampling multiple candidate SQL generations and propose our method, MSc-SQL, to critique them using associated metadata. Our sample critiquing model evaluates multiple outputs simultaneously, achieving state-of-the-art performance compared to other open-source models while remaining competitive with larger models at a much lower cost. Full code can be found at https://github.com/layer6ai-labs/msc-sql.

View on arXivView PDF

Code Implementations(1)

layer6ai-labs/msc-sqlOfficial100%
MIT

Text-2-SQL

192Python, DockerfileOct 16, 20241 years agoMIT

Similar Papers

Open-SQL Framework: Enhancing Text-to-SQL on Open-source Large Language Models

84% match2024

BASE-SQL: A powerful open source Text-To-SQL baseline approach

84% match1 repo132025

SQLfuse: Enhancing Text-to-SQL Performance through Comprehensive LLM Synergy

84% match2024

DTS-SQL: Decomposed Text-to-SQL with Small Large Language Models

84% match2024

CodeS: Towards Building Open-source Language Models for Text-to-SQL

84% match1 repo1992024

SelECT-SQL: Self-correcting ensemble Chain-of-Thought for Text-to-SQL

84% match1 repo52024

Cite this paper

@article{gorti2024mscsql,
  title  = {MSc-SQL: Multi-Sample Critiquing Small Language Models For Text-To-SQL Translation},
  author = {Satya Krishna Gorti and Ilan Gofman and Zhaoyan Liu and Jiapeng Wu and Noël Vouitsis and Guangwei Yu and Jesse C. Cresswell and Rasa Hosseinzadeh},
  year   = {2024},
  eprint = {2410.12916},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2410.12916v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2410.12916.svg)](https://codeofpaper.com/paper/2410.12916)

Discussion