---
dataset_info:
  features:
  - name: question_id
    dtype: int64
  - name: category
    dtype: string
  - name: turns
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: reference
    sequence: string
  - name: history
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: finished
    dtype: bool
  splits:
  - name: train
    num_bytes: 85821
    num_examples: 80
  download_size: 55138
  dataset_size: 85821
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---