README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: ground_truth
      dtype: string
  splits:
    - name: train
      num_bytes: 1522057046.31
      num_examples: 16310
    - name: validation
      num_bytes: 194153760.85
      num_examples: 2025
    - name: test
      num_bytes: 202941033.92
      num_examples: 2185
  download_size: 1884504994
  dataset_size: 1919151841.08
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
language:
  - hu