metadata
dataset_info:
features:
- name: contexts
dtype: string
- name: questions
dtype: string
- name: answers
struct:
- name: answer_start
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 10360301
num_examples: 8308
- name: validation
num_bytes: 3803178
num_examples: 2676
download_size: 1618809
dataset_size: 14163479