---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: detections
    list:
    - name: attributes
      sequence: string
    - name: bbox
      sequence: float64
    - name: labels
      sequence: string
  splits:
  - name: train
    num_bytes: 8645822561.845001
    num_examples: 108077
  download_size: 15667573347
  dataset_size: 8645822561.845001
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---