metadata
dataset_info:
features:
- name: paragraphs
list:
- name: context
dtype: string
- name: qas
list:
- name: answers
list:
- name: answer_start
dtype: int64
- name: text
dtype: string
- name: id
dtype: string
- name: question
dtype: string
- name: title
dtype: string
splits:
- name: train
num_bytes: 21663431
num_examples: 442
download_size: 12035487
dataset_size: 21663431
configs:
- config_name: default
data_files:
- split: train
path: data/train-*