metadata
dataset_info:
features:
- name: image
dtype: image
- name: objects
list:
- name: bbox
sequence: int64
- name: category
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 830056
num_examples: 11
download_size: 826803
dataset_size: 830056
configs:
- config_name: default
data_files:
- split: train
path: data/train-*