Sample More Think Less Gfpo

Group Filtered Policy Optimization (GFPO) reduces inference-time computation by sampling larger groups during training and filtering responses based on length and token efficiency to teach models efficient reasoning.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 2dbf409f7b

Frequently asked questions

npx get-skill add adu2021/sample-more-think-less-gfpo