metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
sequence:
- name: text
dtype: string
- name: answer_start
dtype: int32
splits:
- name: train
num_bytes: 23383729
num_examples: 15036
- name: validation
num_bytes: 3018276
num_examples: 1864
- name: test
num_bytes: 2798098
num_examples: 1910
download_size: 7693225
dataset_size: 29200103
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*