metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: watermark_flag
dtype: bool
splits:
- name: train
num_bytes: 133046619
num_examples: 825
download_size: 133035681
dataset_size: 133046619
configs:
- config_name: default
data_files:
- split: train
path: data/train-*