metadata
dataset_info:
features:
- name: src
dtype: string
- name: tgt
dtype: string
splits:
- name: train
num_bytes: 3880301105
num_examples: 10572313
download_size: 2054879767
dataset_size: 3880301105
configs:
- config_name: default
data_files:
- split: train
path: data/train-*