---
dataset_info:
  features:
  - name: question_id
    dtype: int64
  - name: prompt
    dtype: string
  - name: response_a
    dtype: string
  - name: response_b
    dtype: string
  - name: is_response_0_safe
    dtype: bool
  - name: is_response_1_safe
    dtype: bool
  - name: better_response_id
    dtype: int64
  - name: safer_response_id
    dtype: int64
  - name: better_response
    dtype: string
  - name: safer_response
    dtype: string
  - name: ranked_responses
    sequence: string
  splits:
  - name: train
    num_bytes: 1418163
    num_examples: 1000
  download_size: 736108
  dataset_size: 1418163
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---