metadata
dataset_info:
features:
- name: text_with_holes
dtype: string
- name: text_candidates
dtype: string
- name: A
dtype: string
- name: B
dtype: string
- name: C
dtype: string
- name: D
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 2167491
num_examples: 1482
download_size: 1219458
dataset_size: 2167491
configs:
- config_name: default
data_files:
- split: train
path: data/train-*