README.md

---
dataset_info:
  features:
  - name: image_id
    dtype: int64
  - name: image_path
    dtype: string
  - name: bbox
    sequence: int64
  - name: category
    dtype: string
  - name: category_id
    dtype: int64
  - name: area
    dtype: int64
  - name: iscrowd
    dtype: int64
  splits:
  - name: train
    num_bytes: 54278
    num_examples: 417
  - name: valid
    num_bytes: 15158
    num_examples: 117
  - name: test
    num_bytes: 8698
    num_examples: 67
  download_size: 40799
  dataset_size: 78134
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---