PDC & DM-SFT: A Road for LLM SQL Bug-Fix Enhancing
This addresses SQL bug-fixing for developers using Code LLMs, but it is incremental as it builds on existing models with new training methods.
The paper tackles the problem of SQL bug-fixing in Code LLMs, which often struggle with bug repair despite strong code generation abilities, and introduces PDC and DM-SFT methods that result in models outperforming larger current best-performing models.
Code Large Language Models (Code LLMs), such as Code llama and DeepSeek-Coder, have demonstrated exceptional performance in the code generation tasks. However, most existing models focus on the abilities of generating correct code, but often struggle with bug repair. We introduce a suit of methods to enhance LLM's SQL bug-fixing abilities. The methods are mainly consisted of two parts: A Progressive Dataset Construction (PDC) from scratch and Dynamic Mask Supervised Fine-tuning (DM-SFT). PDC proposes two data expansion methods from the perspectives of breadth first and depth first respectively. DM-SFT introduces an efficient bug-fixing supervised learning approach, which effectively reduce the total training steps and mitigate the "disorientation" in SQL code bug-fixing training. In our evaluation, the code LLM models trained with two methods have exceeds all current best performing model which size is much larger.