Clean Code, Better Models: Enhancing LLM Performance with Smell-Cleaned Dataset

TL;DR

This paper systematically investigates the presence of code smells in LLM training datasets and outputs, proposing an automatic cleaning tool to improve dataset quality and enhance LLM performance in code-related tasks.

Contribution

It introduces SmellCC, an automatic code smell cleaning tool, and demonstrates its effectiveness in improving LLM training data and downstream task performance.

Findings

Code smells are prevalent in benchmark datasets and LLM outputs.

Cleaning code smells improves LLM fine-tuning and downstream task results.

The curated dataset leads to better code quality in generated outputs.

Abstract

The Large Language Models (LLMs) have demonstrated great potential in code-related tasks. However, most research focuses on improving the output quality of LLMs (e.g., correctness), and less attention has been paid to the LLM input (e.g., the training code quality). Given that code smells are widely existed in practice and can negatively impact software maintainability and readability, this study takes the first systematic research to assess and improve dataset quality in terms of code smells. In this work, we first conduct a preliminary study to explore the presence of code smells in a popular benchmark dataset (i.e., CodeSearchNet-Python}) and evaluate the output of several popular LLMs (i.e., DeepSeek-Coder, CodeLlama, and MagiCoder), revealing that code smell issues extensively exist in LLM's input (e.g., benchmark dataset) and output (e.g., generated code). We then conduct our systematic research by taking three main steps: Firstly, we propose an LLM-based code smell cleaning tool, named SmellCC, which automatically refactors and removes code smells. To evaluate the correctness of the code refactoring, we construct a test set of 50 repositories sourced from the CodeSearchNet-Python benchmark for functional testing. Then we apply our curated smell-cleaned dataset to fine-tune two LLMs (i.e., DeepSeek-V2 and Qwen-Coder) to explore their potential for generating high-quality code. Thirdly, we investigate the impact of code smells on two downstream tasks: code completion and code search. Lastly, we derive several actionable implications for software engineering researchers and industry practitioners from our findings.