查询与征服：基于执行引导的SQL生成

Query and Conquer: Execution-Guided SQL Generation

March 31, 2025

作者: Łukasz Borchmann, Marek Wydmuch

cs.AI

摘要

我们提出了一种创新方法，用于生成复杂输出，显著提升了文本到SQL任务的准确性。该方法通过利用执行结果，从多个候选中筛选出语义最一致的查询，使得规模更小、成本效益更高的模型能够超越计算密集型的推理方法，如o1、o3-mini和DeepSeek R1，同时将推理成本降低多达30倍。该方法能够无缝集成到现有模型中，为实现最先进的SQL生成提供了一条实用且可扩展的路径。

English

We propose a novel approach for generating complex outputs that significantly improves accuracy in text-to-SQL tasks. Our method leverages execution results to select the most semantically consistent query from multiple candidates, enabling smaller, cost-effective models to surpass computationally intensive reasoning methods such as o1, o3-mini, and DeepSeek R1 while reducing inference cost by as much as 30 times. It integrates effortlessly with existing models, offering a practical and scalable pathway to state-of-the-art SQL generation.