metadata
dataset_info:
features:
- name: document_id
dtype: large_string
- name: split
dtype: large_string
- name: title
dtype: large_string
- name: kind
dtype: large_string
- name: source
dtype: large_string
- name: text
dtype: large_string
- name: summary
dtype: large_string
- name: question
large_list:
- name: question
dtype: large_string
- name: answers
large_list: large_string
- name: paragraphs
large_list:
- name: paragraph_id
dtype: int64
- name: paragraph
dtype: large_string
splits:
- name: train
num_bytes: 641776490
num_examples: 1102
- name: valid
num_bytes: 66520204
num_examples: 115
- name: test
num_bytes: 193027505
num_examples: 355
download_size: 546634063
dataset_size: 901324199
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*