---
dataset_info:
  features:
  - name: original_text
    dtype: string
  - name: text
    dtype: string
  - name: preference_id
    dtype: string
  - name: topic
    dtype: string
  - name: preference_value
    dtype: string
  - name: rejected_value
    dtype: string
  - name: uid
    dtype: string
  - name: variant
    dtype: string
  splits:
  - name: train
    num_bytes: 4016615
    num_examples: 1596
  download_size: 1676389
  dataset_size: 4016615
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---