metadata
dataset_info:
features:
- name: en
dtype: string
- name: br
dtype: string
- name: en_image_clean
dtype: image
- name: br_image_clean
dtype: image
- name: en_image_aug
dtype: image
- name: br_image_aug
dtype: image
- name: metadata
dtype: string
splits:
- name: train
num_bytes: 8805011749
num_examples: 100000
download_size: 8562511938
dataset_size: 8805011749
configs:
- config_name: default
data_files:
- split: train
path: data/train-*