metadata
dataset_info:
features:
- name: image
dtype: image
- name: normalized_bboxes
sequence:
- name: name
dtype: string
- name: bbox
sequence: string
splits:
- name: train
num_bytes: 94167197
num_examples: 1001
download_size: 1182357
dataset_size: 94167197
configs:
- config_name: default
data_files:
- split: train
path: data/train-*