metadata
dataset_info:
features:
- name: answers
struct:
- name: answer_start
sequence: int64
- name: text
sequence: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: id
dtype: string
splits:
- name: train
num_bytes: 13071626
num_examples: 38783
- name: validation
num_bytes: 3262468
num_examples: 9697
download_size: 6565635
dataset_size: 16334094