无需容器的轻量RL训练框架，显著降低资源消耗。
SWE-MiniSandbox: Container-Free Reinforcement Learning for Building Software Engineering Agents
- 用内核级隔离机制替代容器，实现任务独立运行
- 磁盘占用降至容器方案的5%，环境准备时间缩短至25%
- 适合资源受限环境下的软件工程智能体规模化训练
强化学习（RL）已成为训练软件工程（SWE）智能体的关键范式，但现有流程通常依赖每个任务的独立容器以保证隔离性。在大规模场景下，预构建的容器镜像会带来巨大的存储开销、环境初始化缓慢，并需容器管理权限。我们提出 SWE-MiniSandbox，一种轻量级、无需容器的方法，可在不牺牲隔离性的前提下实现 SWE 智能体的可扩展强化学习训练。该方法不依赖实例级容器，而是利用内核级机制为每个任务提供隔离的工作空间，并结合轻量级环境预缓存技术，彻底消除对庞大容器镜像的需求。实验表明，该方法将磁盘使用量降低至容器基线的约5%，环境准备时间减少至约25%。实证结果证明，SWE-MiniSandbox 在评估性能上与标准容器化方案相当。通过移除对重型容器基础设施的依赖，SWE-MiniSandbox 为资源受限的研究环境提供了实用且易部署的 RL 基础架构，推动 SWE 智能体的规模化发展。
原文摘要 · Abstract (English)
Reinforcement learning (RL) has become a key paradigm for training software engineering (SWE) agents, but existing pipelines typically rely on per-task containers for isolation. At scale, pre-built container images incur substantial storage overhead, slow environment setup, and require container-management privileges. We propose SWE-MiniSandbox, a lightweight, container-free method that enables scalable RL training of SWE agents without sacrificing isolation. Instead of relying on per-instance containers, SWE-MiniSandbox executes each task in an isolated workspace backed by kernel-level mechanisms, substantially reducing system overhead. It leverages lightweight environment pre-caching techniques to eliminate the need for bulky container images. As a result, our approach lowers disk usage to approximately 5\% of that required by container-based pipelines and reduces environment preparation time to about 25\% of the container baseline. Empirical results demonstrate that SWE-MiniSandbox achieves evaluation performance comparable to standard container-based pipelines. By removing the dependency on heavy container infrastructure, SWE-MiniSandbox offers a practical and accessible foundation for scaling RL-based SWE agents, particularly in resource-constrained research environments.
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。