Optimizing Small Language Models for NL2SQL via Chain-of-Thought Fine-Tuning

ArXi:2603.22942v1 Announce Type: new Translating Natural Language to SQL (NL2SQL) remains a critical bottleneck for cratization of data in enterprises. Although Large Language Models (LLMs) like Gemini 2.5 and other LLMs have nstrated impressive zero-shot capabilities, their high inference costs limit deployment at scale. This paper explores the efficacy of fine-tuning both large and small language models on NL2SQL tasks. Our research reveals a counter-intuitive scaling phenomenon.