---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: annotation
    dtype: image
  - name: inst_ids
    list: int32
  - name: sem_ids
    list: int32
  splits:
  - name: train
    num_bytes: 808680849.0
    num_examples: 384
  - name: validation
    num_bytes: 75394106.0
    num_examples: 35
  download_size: 226719153
  dataset_size: 884074955.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---