---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: modality
    dtype: string
  splits:
  - name: train
    num_bytes: 1723286375.4336302
    num_examples: 9000
  - name: test
    num_bytes: 108794824.71891744
    num_examples: 3000
  download_size: 1682256345
  dataset_size: 1832081200.1525476
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---