metadata
license: mit
datasets:
- OpenAssistant/oasst1
tags:
- RM
The model trained with L=1 (Length loss weight) and O=1 (Orthogonal loss weight).
license: mit
datasets:
- OpenAssistant/oasst1
tags:
- RM
The model trained with L=1 (Length loss weight) and O=1 (Orthogonal loss weight).