README.md

metadata

datasets:
  - trl-lib/ultrafeedback_binarized
base_model:
  - ComparisonPO/Llama-3.2-3B-Instruct-DPO