metadata
dataset_info:
features:
- name: image
dtype: image
- name: boxes
sequence:
sequence: int32
length: 4
- name: labels
sequence: string
splits:
- name: train
num_bytes: 52250258
num_examples: 48
download_size: 52254231
dataset_size: 52250258
configs:
- config_name: default
data_files:
- split: train
path: data/train-*