metadata
dataset_info:
features:
- name: context
dtype: string
- name: qas
list:
- name: answer
dtype: string
- name: question
dtype: string
- name: subset
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1395598661
num_examples: 664062
download_size: 733926862
dataset_size: 1395598661
configs:
- config_name: default
data_files:
- split: train
path: data/train-*