---
dataset_info:
  features:
  - name: query
    dtype: string
  - name: images
    sequence: string
  - name: response
    dtype: string
  splits:
  - name: train
    num_bytes: 19362522
    num_examples: 91000
  download_size: 4806185
  dataset_size: 19362522
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---