README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: text
      sequence: string
    - name: index
      dtype: int64
  splits:
    - name: train
      num_bytes: 51733249508.27
      num_examples: 81355
  download_size: 51732268420
  dataset_size: 51733249508.27
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: apache-2.0
task_categories:
  - image-to-text
language:
  - en
size_categories:
  - 10K<n<100K

WikiArt Dataset captioned using vikhyatk/moondream2 model with prompt : Generate a short, simple and only visually descriptive caption for this image.