InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct

Recent advancements in open-source code large language models (LLMs) have been driven by fine-tuning on the data generated from powerful closed-source LLMs, which are expensive to obtain. This paper explores whether it is possible to use a fine-tuned open-source model to generate additional data to augment its instruction-tuning dataset. We make two observations: (1) A code snippet can serve as the response to different instructions. (2) Instruction-tuned code LLMs perform better at translating code into instructions than the reverse. Based on these observations, we propose Inverse-Instruct, a data augmentation technique that uses a fine-tuned LLM to generate additional instructions of code responses from its own training dataset. The additional instruction-response pairs are added to the original dataset, and a stronger code LLM can be obtained by fine-tuning on the augmented dataset. We empirically validate Inverse-Instruct on a range of open-source code models (e.g. CodeLlama-Python and DeepSeek-Coder) and benchmarks (e.g., HumanEval(+), MBPP(+), DS-1000 and MultiPL-E), showing it consistently improves the base models.

Authors

• Yutong Wu Duke University
• Di Huang Beihang University
• Wenxuan Shi SKL of Processors, Institute of Computing Technology, CAS University of Chinese Academy of Sciences
• Wei Wang
• Yewen Pu Autodesk Research
• Lingzhe Gao Baidu Inc., Beijing, China
• Shihao Liu Baidu Inc., Beijing, China
• Ziyuan Nan SKL of Processors, Institute of Computing Technology, CAS University of Chinese Academy of Sciences
• Kaizhao Yuan SKL of Processors, Institute of Computing Technology, CAS University of Chinese Academy of Sciences
• Rui Zhang
• Xishan Zhang SKL of Processors, Institute of Computing Technology, CAS
• Zidong Du SKL of Processors, Institute of Computing Technology, CAS
• Qi Guo SKL of Processors, Institute of Computing Technology, CAS
• Dawei Yin
• Xing Hu SKL of Processors, Institute of Computing Technology, CAS
• Yunji Chen SKL of Processors, Institute of Computing Technology, CAS University of Chinese Academy of Sciences

Keywords

No keywords are indexed for this paper.