metadata
dataset_info:
features:
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
struct:
- name: answer_start
sequence: int64
- name: text
sequence: string
- name: paragraph_id
dtype: string
splits:
- name: train
num_bytes: 74747951
num_examples: 75722
- name: validation
num_bytes: 11404322
num_examples: 10570
- name: test
num_bytes: 10925430
num_examples: 11877
download_size: 17426212
dataset_size: 97077703
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*