---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 4215984699.0
    num_examples: 12000
  - name: validation
    num_bytes: 126870900.0
    num_examples: 400
  download_size: 4504169676
  dataset_size: 4342855599.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---