Dubo-SQL：用于文本到SQL的多样化检索增强生成与微调
Dubo-SQL: Diverse Retrieval-Augmented Generation and Fine Tuning for Text-to-SQL
- Mercator Technologies（墨卡托科技）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
Dubo-SQL通过低成本微调、多样化RAG及新输入输出格式，在BIRD-SQL上以GPT-3.5 Turbo超越GPT-4模型，并推出v2进一步提升性能。
AI中文摘要：
当前自动化文本到SQL（text-to-SQL）的最先进技术（SOTA）在BIRD-SQL基准上的执行准确率（EX）仍远低于人类专家水平。最准确的方法也既慢又昂贵。为了在降低成本和提高速度的同时推进文本到SQL的SOTA，我们探索了低成本微调、新颖的多样化检索增强生成（RAG）方法以及有助于大型语言模型（LLMs）实现更高EX的新输入和输出格式的组合。我们引入了两种新方法，Dubo-SQL v1和v2。Dubo-SQL v1在BIRD-SQL的保留测试集上创下了新的EX记录。Dubo-SQL v2在BIRD-SQL开发集上取得了更高的性能。Dubo-SQL v1依赖于OpenAI的LLMs，但使用低成本的GPT-3.5 Turbo，同时超过了使用更昂贵的GPT-4的次优OpenAI模型的性能。Dubo-SQL v1比使用GPT-3.5的次优模型性能高出20%以上。Dubo-SQL v2使用GPT-4 Turbo和RAG代替微调来进一步提升EX。
英文摘要：
The current state-of-the-art (SOTA) for automated text-to-SQL still falls well short of expert human performance as measured by execution accuracy (EX) on the BIRD-SQL benchmark. The most accurate methods are also slow and expensive. To advance the SOTA for text-to-SQL while reducing cost and improving speed, we explore the combination of low-cost fine tuning, novel methods for diverse retrieval-augmented generation (RAG) and new input and output formats that help large language models (LLMs) achieve higher EX. We introduce two new methods, Dubo-SQL v1 and v2. Dubo-SQL v1 sets a new record for EX on the holdout test set of BIRD-SQL. Dubo-SQL v2 achieves even higher performance on the BIRD-SQL dev set. Dubo-SQL v1 relies on LLMs from OpenAI, but uses the low-cost GPT-3.5 Turbo while exceeding the performance of the next-best model using OpenAI, which instead uses the more expensive GPT-4. Dubo-SQL v1 exceeds the performance of the next-best model using GPT-3.5 by over 20%. Dubo-SQL v2 uses GPT-4 Turbo and RAG in place of fine tuning to push EX higher.
