metadata
dataset_info:
features:
- name: raw_json
dtype: string
- name: images
list: image
splits:
- name: train
num_bytes: 4841348326.712
num_examples: 9927
download_size: 486961149
dataset_size: 4841348326.712
configs:
- config_name: default
data_files:
- split: train
path: data/train-*