用低秩适配融合多编码器，提升代码注释分类准确率
LoRA-MME: Multi-Model Ensemble of LoRA-Tuned Encoders for Code Comment Classification
- 采用LoRA分别微调四个编码器，再加权融合预测结果
- 测试集上加权F1达0.7906，宏平均F1为0.6867
- 适合追求高精度但可接受较高推理开销的代码分析场景
代码注释分类是自动化软件文档与分析的关键任务。在NLBSE'26工具竞赛中，我们提出LoRA-MME，一种基于参数高效微调（PEFT）的多模型集成架构。该方法通过独立使用低秩适配（LoRA）微调UniXcoder、CodeBERT、GraphCodeBERT和CodeBERTa四类不同Transformer编码器，并采用可学习的加权集成策略聚合其预测结果，实现跨Java、Python和Pharo的多标签分类性能最大化，同时避免全模型微调带来的内存开销。在测试集上，该工具取得0.7906的加权F1和0.6867的宏平均F1。然而，集成结构带来的计算成本导致最终提交得分为41.20%，凸显了语义准确性与推理效率之间的权衡。
原文摘要 · Abstract (English)
Code comment classification is a critical task for automated software documentation and analysis. In the context of the NLBSE'26 Tool Competition, we present LoRA-MME, a Multi-Model Ensemble architecture utilizing Parameter-Efficient Fine-Tuning (PEFT). Our approach addresses the multi-label classification challenge across Java, Python, and Pharo by combining the strengths of four distinct transformer encoders: UniXcoder, CodeBERT, GraphCodeBERT, and CodeBERTa. By independently fine-tuning these models using Low-Rank Adaptation(LoRA) and aggregating their predictions via a learned weighted ensemble strategy, we maximize classification performance without the memory overhead of full model fine-tuning. Our tool achieved an F1 Weighted score of 0.7906 and a Macro F1 of 0.6867 on the test set. However, the computational cost of the ensemble resulted in a final submission score of 41.20%, highlighting the trade-off between semantic accuracy and inference efficiency.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。