---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: prompt
    dtype: string
  - name: predicted_label
    dtype: string
  - name: confidence
    dtype: float64
  - name: safe_prob
    dtype: float64
  - name: violence_prob
    dtype: float64
  - name: sexual_prob
    dtype: float64
  - name: gore_prob
    dtype: float64
  - name: safe
    dtype: int64
  - name: violence
    dtype: int64
  - name: sexual
    dtype: int64
  - name: gore
    dtype: int64
  splits:
  - name: train
    num_bytes: 359831425
    num_examples: 1000
  download_size: 359807204
  dataset_size: 359831425
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---