---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_id
    dtype: int64
  - name: size
    sequence: int64
  - name: proportion
    dtype: float64
  - name: answers
    sequence: string
  splits:
  - name: train
    num_bytes: 1396533.0
    num_examples: 52
  download_size: 1396614
  dataset_size: 1396533.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---