metadata
dataset_info:
features:
- name: document
dtype: string
- name: summary
dtype: string
- name: isAnswer
dtype: int64
- name: q_id
dtype: string
- name: sample_id
dtype: string
- name: isMultisent
dtype: bool
splits:
- name: train
num_bytes: 13661601
num_examples: 32374
download_size: 1669589
dataset_size: 13661601
configs:
- config_name: default
data_files:
- split: train
path: data/train-*