Code Comparison Tuning for Code Large Language Models
This work addresses bug-fixing capabilities for developers using code LLMs, representing an incremental improvement over existing tuning methods.
The paper tackles the problem of subtle code errors in code large language models by introducing Code Comparison Tuning (CCT), which integrates token- and sequence-level comparisons into instruction tuning, resulting in up to a 4-point improvement in pass@1 scores on the HumanEvalFix benchmark.
We present Code Comparison Tuning (CCT), a simple and effective tuning method for code large language models (Code LLMs) to better handle subtle code errors. Specifically, we integrate the concept of comparison into instruction tuning, both at the token and sequence levels, enabling the model to discern even the slightest deviations in code. To compare the original code with an erroneous version containing manually added code errors, we use token-level preference loss for detailed token-level comparisons. Additionally, we combine code segments to create a new instruction tuning sample for sequence-level comparisons, enhancing the model's bug-fixing capability. Experimental results on the HumanEvalFix benchmark show that CCT surpasses instruction tuning in pass@1 scores by up to 4 points across diverse code LLMs, and extensive analysis demonstrates the effectiveness of our method.