metadata
dataset_info:
features:
- name: text
dtype: string
- name: source
dtype: string
- name: num_tokens
dtype: int64
splits:
- name: train
num_bytes: 2287700
num_examples: 20904
download_size: 923632
dataset_size: 2287700
configs:
- config_name: default
data_files:
- split: train
path: data/train-*