metadata
dataset_info:
features:
- name: text
dtype: string
- name: mask
sequence: int64
splits:
- name: train
num_bytes: 435927
num_examples: 5000
download_size: 164074
dataset_size: 435927
configs:
- config_name: default
data_files:
- split: train
path: data/train-*