metadata
dataset_info:
features:
- name: twi
dtype: string
- name: en
dtype: string
- name: col_data
sequence: string
- name: direction
dtype: string
splits:
- name: train
num_bytes: 5256867
num_examples: 3337
download_size: 1805598
dataset_size: 5256867
configs:
- config_name: default
data_files:
- split: train
path: data/train-*