metadata
dataset_info:
features:
- name: Q
dtype: string
- name: A
dtype: string
- name: D_star
sequence: string
- name: D_i
sequence: string
splits:
- name: train
num_bytes: 368275577
num_examples: 42926
download_size: 113636800
dataset_size: 368275577
configs:
- config_name: default
data_files:
- split: train
path: data/train-*