---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: message
    dtype: string
  - name: answer
    dtype: string
  - name: task_type
    dtype: string
  - name: dataset_name
    dtype: string
  - name: meta_info
    dtype: string
  splits:
  - name: train
    num_bytes: 87910299.0
    num_examples: 1600
  download_size: 86188990
  dataset_size: 87910299.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---