---
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
---