by SWE-bench
Open source · 93k downloads · 6 likes
The SWE-agent LM 7B model is a language model specialized in software engineering, designed to assist autonomous agents in resolving code development and maintenance tasks. Trained on trajectories generated from interactions between SWE-agent and advanced models like Claude 3.7 Sonnet, it excels in understanding and generating code, as well as executing complex operations on software repositories. Its key capabilities include analyzing technical issues, proposing fixes, and automating development workflows, making it a valuable tool for teams seeking to accelerate debugging and continuous improvement processes. This model stands out for its open-source approach and seamless integration with the SWE-agent ecosystem, offering an accessible and high-performance solution for developers and AI researchers.
SWE-agent-LM-7B is a Language Model for Software Engineering trained using the SWE-smith toolkit. We introduce this model as part of our work: SWE-smith: Scaling Data for Software Engineering Agents.
SWE-agent-LM-7B is 100% open source. Training this model was simple - we fine-tuned Qwen 2.5 Coder Instruct on 5k trajectories generated by SWE-agent + Claude 3.7 Sonnet. The dataset can be found here.
SWE-agent-LM-7B is compatible with SWE-agent. Running this model locally only takes a few steps! Check here for more instructions on how to do so.
If you found this work exciting and want to push SWE-agents further, please feel free to connect with us (the SWE-bench team) more!