---
dataset_info:
  features:
  - name: messages
    dtype: string
  - name: image
    dtype: image
  - name: answer
    dtype: string
  - name: meta_info
    struct:
    - name: dataset
      dtype: string
    - name: name
      dtype: string
    - name: question_type
      dtype: string
  splits:
  - name: train
    num_bytes: 2380197.0
    num_examples: 400
  download_size: 2212030
  dataset_size: 2380197.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---