metadata
dataset_info:
features:
- name: raw
dtype: string
- name: trans
dtype: string
- name: source_lang
dtype: string
- name: target_lang
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 16892187
num_examples: 82589
- name: validation
num_bytes: 2112277
num_examples: 10334
download_size: 11528173
dataset_size: 19004464
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*