How to Fine-Tune a Reasoning Model? A Teacher-Student Cooperation Framework to Synthesize Student-Consistent SFT Data

Zixian Huang, Kaichen Yang, xuhuang, Feiyang Hao, Qiming Ge, Bowen Li, He Du, Kai Chen, Qipeng Guo

How to Fine-Tune a Reasoning Model? A Teacher-Student Cooperation Framework to Synthesize Student-Consistent SFT Data: 35 upvotes on Hugging Face Daily Papers, #3 of 29 papers on 2026-04-17. Day-by-day upvote history.

A widely adopted strategy for model enhancement is to use synthetic data generated by a stronger model for supervised fine-tuning (SFT). However, for emerging reasoning models like Qwen3-8B, this approach often fails to improve reasoning capabilities and can even lead to a substantial drop in performance. In this work, we identify substantial stylistic divergence between teacher generated data and the distribution of student as a major factor impacting SFT. To bridge this gap, we propose a Teacher-Student Cooperation Data Synthesis framework (TESSY), which interleaves teacher and student models to alternately generate style and non-style tokens. Consequently, TESSY produces synthetic sequences that inherit the advanced reasoning capabilities of the teacher while maintaining stylistic consistency with the distribution of the student. In experiments on code generation using GPT-OSS-120B as the teacher, fine-tuning Qwen3-8B on teacher-generated data leads to performance drops of 3.25% on LiveCodeBench-Pro and 10.02% on OJBench, whereas TESSY achieves improvements of 11.25% and 6.68%.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.