metadata
dataset_info:
features:
- name: src_lang
dtype: int64
- name: tgt_lang
dtype: int64
- name: src_text
dtype: string
- name: tgt_text
dtype: string
- name: score
dtype: float64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1049666688
num_examples: 1678579
download_size: 554320464
dataset_size: 1049666688
configs:
- config_name: default
data_files:
- split: train
path: data/train-*