metadata
dataset_info:
features:
- name: source
dtype: string
- name: target
dtype: string
splits:
- name: train
num_bytes: 5967026
num_examples: 20106
download_size: 3715296
dataset_size: 5967026
configs:
- config_name: default
data_files:
- split: train
path: data/train-*