Open Models Almanac
Methodspost-trainingsupervised fine-tuning

single mention · post-training

Evaluation-based early stopping during SFT

Stops SFT based on evaluation scores rather than completing a fixed training schedule.

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 early stopping based on evaluation scores.

Filed under post-training › supervised fine-tuning