Datasets:

nu-dialogue
/

sfcoco2022

dataset_info:
  features:
    - name: image
      dtype: image
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 366416649.50223213
      num_examples: 806
    - name: test
      num_bytes: 41865941.49776786
      num_examples: 90
  download_size: 405465686
  dataset_size: 408282591
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
task_categories:
  - image-to-text
language:
  - ja