metadata
dataset_info:
features:
- name: translation
struct:
- name: de
dtype: string
- name: en
dtype: string
splits:
- name: train
num_bytes: 187718795
num_examples: 576000
- name: validation
num_bytes: 736028
num_examples: 3000
- name: test
num_bytes: 766158
num_examples: 3003
download_size: 107108210
dataset_size: 189220981
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*