metadata
dataset_info:
features:
- name: text
dtype: string
- name: language
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 104504087
num_examples: 3994
download_size: 55336521
dataset_size: 104504087
configs:
- config_name: default
data_files:
- split: train
path: data/train-*