metadata
dataset_info:
features:
- name: text
dtype: string
- name: neighbor
dtype: string
- name: token
dtype: string
- name: position
dtype: int64
splits:
- name: train
num_bytes: 9472340
num_examples: 10719
download_size: 775073
dataset_size: 9472340
configs:
- config_name: default
data_files:
- split: train
path: data/train-*