metadata
dataset_info:
features:
- name: image
dtype: image
splits:
- name: train
num_bytes: 888141950.333
num_examples: 4637
download_size: 897984928
dataset_size: 888141950.333
configs:
- config_name: default
data_files:
- split: train
path: data/train-*