metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 32200002
num_examples: 900000
- name: test
num_bytes: 3577778
num_examples: 100000
download_size: 18076399
dataset_size: 35777780
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*