OpenLMLab

Secrets of RLHF in Large Language Models Part I: PPO

LOMO: LOw-Memory Optimization

GAOKAO-Bench is an evaluation framework that utilizes GAOKAO questions as a dataset to evaluate large language models.