RLHFlow

icon indicating an email [email protected]

icon location United States of America icon location Code for the Workflow of Reinforcement Learning from Human Feedback (RLHF)

Results 3 repositories owned by RLHFlow

Directional-Preference-Alignment

45
Stars
2
Forks
Watchers

Directional Preference Alignment

verified publisher icon RLHFlow

RLHF-Reward-Modeling

1.5k
Stars
103
Forks
1.5k
Watchers

Recipes to train reward model for RLHF.

verified publisher icon RLHFlow

Online-RLHF

536
Stars
49
Forks
536
Watchers

A recipe for online RLHF and online iterative DPO.

verified publisher icon RLHFlow