metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: text_en
dtype: string
- name: text_id
dtype: string
- name: target_embedding
sequence: float32
- name: input_ids_en
sequence: int64
- name: attention_mask_en
sequence: int64
- name: input_ids_id
sequence: int64
- name: attention_mask_id
sequence: int64
splits:
- name: train
num_bytes: 11676096944
num_examples: 1000000
download_size: 4112187708
dataset_size: 11676096944