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: 2646337.9600498625
num_examples: 2847
- name: validation
num_bytes: 4913.35830947971
num_examples: 5
download_size: 4265888
dataset_size: 2651251.3183593424