metadata
dataset_info:
features:
- name: question_text
dtype: string
- name: document_title
dtype: string
- name: language
dtype: string
- name: annotations
struct:
- name: answer_start
sequence: int64
- name: answer_text
sequence: string
- name: document_plaintext
dtype: string
- name: document_url
dtype: string
splits:
- name: train
num_bytes: 32084629.326371837
num_examples: 29868
- name: validation
num_bytes: 3778385.324427767
num_examples: 3712
download_size: 16354337
dataset_size: 35863014.6507996