single source · post-training
Light supervised fine-tuning on teacher-distribution data
Lightly fine-tunes a student model on data drawn from a teacher's training distribution before reinforcement learning.
The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See supervised fine-tuning for the methods filed beside it.
Also called teacher-distribution SFT warmup.
Filed under post-training › supervised fine-tuning