metadata
dataset_info:
features:
- name: id
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 207119705492.8
num_examples: 3016640
download_size: 206342349838
dataset_size: 207119705492.8
configs:
- config_name: default
data_files:
- split: train
path: data/train-*