metadata
dataset_info:
features:
- name: sentence
dtype: string
- name: option1
dtype: string
- name: option2
dtype: string
- name: answer
dtype: string
- name: qas_id
dtype: int64
splits:
- name: train
num_bytes: 1463187
num_examples: 9248
- name: test
num_bytes: 276255
num_examples: 1767
- name: valid
num_bytes: 199703
num_examples: 1267
download_size: 991176
dataset_size: 1939145
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*