SWE-Prime: Fewer Trajectories, Better Performance
[Inżynieria oprogramowania] Autorzy: Dewu Zheng, Ruizhe Ye, Yanlin Wang i 7 więcej
To improve large language models' ability to resolve real-world software issues, prior work has focused on constructing large-scale agent trajectory datasets and performing supervised fine-tuning (SFT) on successful trajectories. However, task success does not guarantee high-quality supervision: successful trajectories may still contain ineffective, redundant, or risky steps. Directly using such trajectories for SFT can introduce noisy supervision and encourage models to imitate undesirable problem-solving behaviors. Therefore, we propose SWE-Prime, a multi-granularity, two-stage SFT data selection method that progressively filters training data at the trajectory and segment levels. Specifically, the first stage performs trajectory-level screening based on process quality, result quality, and data representativeness, selecting a high-quality and representative subset of successful traje…
- Kategoria arxiv
- cs.SE
- Kategoria
- Inżynieria oprogramowania