metadata
dataset_info:
features:
- name: question
dtype: string
- name: answers
list:
- name: answer_start
dtype: int64
- name: text
dtype: string
- name: answer
dtype: string
- name: context
dtype: string
- name: title
dtype: string
splits:
- name: train
num_bytes: 84600671
num_examples: 25735
download_size: 6476264
dataset_size: 84600671