---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 2987962172.136
    num_examples: 32052
  - name: validation
    num_bytes: 362401360.28
    num_examples: 3888
  - name: test
    num_bytes: 387159994.503
    num_examples: 4061
  download_size: 3647464913
  dataset_size: 3737523526.9190006
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---