---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: image
    dtype: image
  - name: qid
    dtype: string
  - name: prompt
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 5134742331.994
    num_examples: 33203
  download_size: 822711929
  dataset_size: 5134742331.994
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---