metadata
dataset_info:
features:
- name: image
dtype: image
- name: id
dtype: int64
- name: bboxes
sequence:
sequence: int64
- name: labels
sequence: string
- name: category_ids
sequence: int64
splits:
- name: train
num_bytes: 200863470.75
num_examples: 1906
download_size: 198577309
dataset_size: 200863470.75
configs:
- config_name: default
data_files:
- split: train
path: data/train-*