metadata
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-*