metadata
dataset_info:
features:
- name: modified_images
dtype: image
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 2117695764
num_examples: 800
- name: test
num_bytes: 513608173
num_examples: 200
download_size: 2631068796
dataset_size: 2631303937
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*