metadata
dataset_info:
features:
- name: image
dtype: image
- name: annotations
list:
- name: category_id
dtype: int64
- name: bbox
sequence: float64
- name: keypoints
sequence: int64
- name: image_id
dtype: int64
- name: id
dtype: int64
splits:
- name: train
num_bytes: 8811815407.375
num_examples: 64115
- name: validation
num_bytes: 359349066.13
num_examples: 2693
download_size: 8804604126
dataset_size: 9171164473.505
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*