metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 1047958
num_examples: 14000
- name: dev
num_bytes: 150348
num_examples: 2000
- name: test
num_bytes: 298101
num_examples: 4000
download_size: 787307
dataset_size: 1496407