---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: image_id
    dtype: string
  - name: original_answer
    dtype: string
  - name: consensus_ratio
    dtype: float64
  - name: total_responses
    dtype: int64
  splits:
  - name: train
    num_bytes: 6588416.0
    num_examples: 25
  download_size: 6590694
  dataset_size: 6588416.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---