metadata
dataset_info:
features:
- name: image
dtype: image
- name: objects
struct:
- name: bbox
sequence:
sequence:
sequence: int64
- name: categories
sequence: int64
- name: taxonomy_id
sequence: string
- name: scientific_name
sequence: string
- name: boxes
sequence:
sequence: int64
splits:
- name: train
num_bytes: 8101528
num_examples: 82
download_size: 7999668
dataset_size: 8101528
configs:
- config_name: default
data_files:
- split: train
path: data/train-*