---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: response_a
    dtype: string
  - name: response_b
    dtype: string
  - name: more_reasonable
    dtype: string
  splits:
  - name: train
    num_bytes: 1716176
    num_examples: 4808
  download_size: 455224
  dataset_size: 1716176
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---