README.md

metadata

language:
  - en
  - hu
task_categories:
  - image-to-text
dataset_info:
  features:
    - name: img
      dtype: image
    - name: text_EN
      dtype: string
  splits:
    - name: train
      num_bytes: 2034111254
      num_examples: 20000
  download_size: 2033623775
  dataset_size: 2034111254
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*