metadata
dataset_info:
features:
- name: id
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
struct:
- name: answer_start
sequence: int32
- name: text
sequence: string
- name: subset
dtype: string
splits:
- name: train
num_bytes: 78286119
num_examples: 87599
- name: validation
num_bytes: 210177136
num_examples: 163368
- name: test
num_bytes: 210177136
num_examples: 163368
download_size: 217966658
dataset_size: 498640391
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*