metadata
dataset_info:
features:
- name: type
dtype: string
- name: content_without_mask
dtype: string
- name: content_masked
dtype: string
- name: label
dtype: string
- name: options
sequence: string
- name: correct
dtype: string
splits:
- name: train
num_bytes: 19401232
num_examples: 2815
download_size: 7169339
dataset_size: 19401232
configs:
- config_name: default
data_files:
- split: train
path: data/train-*