Datasets:

MosRat
/

gex_dataset_merged

dataset_info:
  features:
    - name: image
      dtype: image
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 16284500676
      num_examples: 1472030
  download_size: 15873671982
  dataset_size: 16284500676
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: mit
task_categories:
  - image-to-text
language:
  - zh
  - en
pretty_name: a
size_categories:
  - 1M<n<10M