---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: prompt
    dtype: string
  - name: relation
    dtype: string
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 785331811
    num_examples: 5057
  - name: validation
    num_bytes: 88607504
    num_examples: 564
  download_size: 866253316
  dataset_size: 873939315
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---