---
dataset_info:
  features:
  - name: index
    dtype: int64
  - name: source_dataset
    dtype: string
  - name: base_answer
    dtype: string
  - name: caption
    dtype: string
  - name: image
    dtype: image
  - name: base_question
    dtype: string
  - name: id
    dtype: string
  - name: responses
    dtype: string
  splits:
  - name: train
    num_bytes: 42334302.0
    num_examples: 1000
  download_size: 41146293
  dataset_size: 42334302.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---