metadata
dataset_info:
features:
- name: image_id
dtype: int64
- name: depth
sequence:
sequence: uint8
splits:
- name: train
num_bytes: 16062050708
num_examples: 74942
download_size: 3956393794
dataset_size: 16062050708
configs:
- config_name: default
data_files:
- split: train
path: data/train-*