metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
sequence: string
- name: context
sequence: string
splits:
- name: train
num_bytes: 12077616
num_examples: 3621
download_size: 6235907
dataset_size: 12077616
configs:
- config_name: default
data_files:
- split: train
path: data/train-*