通过后训练框架提升代码生成智能体的长期任务解决能力。
SWE-Master: Unleashing the Potential of Software Engineering Agents via Post-Training
- 构建全流程后训练框架，融合轨迹合成与真实执行反馈优化
- 在SWE-bench Verified上达61.4%解决率，引入测试时扩展后达70.8%
- 开源透明，适合研究可复现的软件工程智能体
本文提出SWE-Master，一个开源且完全可复现的后训练框架，用于构建高效的软件工程智能体。该框架系统化探索了智能体开发全流程，包括教师轨迹合成与数据整理、长时序监督微调（SFT）、基于真实执行反馈的强化学习（RL），以及推理框架设计。从初始代码能力有限的开源基础模型出发，SWE-Master展示了系统性优化方法如何激发强长时序软件工程任务求解能力。在标准基准SWE-bench Verified上，相同实验条件下，采用Qwen2.5-Coder-32B模型时，解决率达61.4%，显著优于现有开源基线。进一步结合基于大语言模型的环境反馈进行测试时扩展（TTS）后，性能提升至70.8%（TTS@8）。SWE-Master为推进可复现的软件工程智能体研究提供了实用且透明的基础。代码已开源：https://github.com/RUCAIBox/SWE-Master。
原文摘要 · Abstract (English)
In this technical report, we present SWE-Master, an open-source and fully reproducible post-training framework for building effective software engineering agents. SWE-Master systematically explores the complete agent development pipeline, including teacher-trajectory synthesis and data curation, long-horizon SFT, RL with real execution feedback, and inference framework design. Starting from an open-source base model with limited initial SWE capability, SWE-Master demonstrates how systematical optimization method can elicit strong long-horizon SWE task solving abilities. We evaluate SWE-Master on SWE-bench Verified, a standard benchmark for realistic software engineering tasks. Under identical experimental settings, our approach achieves a resolve rate of 61.4\% with Qwen2.5-Coder-32B, substantially outperforming existing open-source baselines. By further incorporating test-time scaling~(TTS) with LLM-based environment feedback, SWE-Master reaches 70.8\% at TTS@8, demonstrating a strong performance potential. SWE-Master provides a practical and transparent foundation for advancing reproducible research on software engineering agents. The code is available at https://github.com/RUCAIBox/SWE-Master.
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。