---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: response_A
    dtype: string
  - name: response_B
    dtype: string
  - name: reasoning
    dtype: string
  - name: preferred
    dtype: string
  splits:
  - name: train
    num_bytes: 475365
    num_examples: 200
  download_size: 301526
  dataset_size: 475365
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---