---
dataset_info:
  features:
  - name: original_prompt
    dtype: string
  - name: original_response
    dtype: string
  - name: prompt_index
    dtype: int64
  - name: section_index
    dtype: int64
  - name: propensity
    dtype: string
  - name: propensity_index
    dtype: int64
  - name: propensity_round
    dtype: int64
  - name: sys_prompt_modified_response
    dtype: string
  - name: grade_sys_prompt_follow
    dtype: string
  - name: propensity_number
    dtype: int64
  - name: sample_type
    dtype: string
  splits:
  - name: train
    num_bytes: 2763029
    num_examples: 1250
  download_size: 1459772
  dataset_size: 2763029
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---