---
dataset_info:
  features:
  - name: prompt_id
    dtype: string
  - name: text
    dtype: string
  - name: hypo_image
    dtype: image
  - name: adversarial_image
    dtype: image
  - name: __index_level_0__
    dtype: int64
  splits:
  - name: train
    num_bytes: 2627656060.32
    num_examples: 1560
  download_size: 2579298148
  dataset_size: 2627656060.32
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---