metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': all_bboxes
'1': relabeled_bboxes
'2': suspect_bboxes
- name: filename
dtype: string
splits:
- name: train
num_bytes: 202947794.916
num_examples: 2649
download_size: 187671396
dataset_size: 202947794.916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*