---
dataset_info:
  features:
  - name: id
    dtype: int32
  - name: language
    dtype: string
  - name: compound
    dtype: string
  - name: sentence_type
    dtype: string
  - name: sentence
    dtype: string
  - name: style
    dtype: string
  - name: correct_image
    dtype: int32
  - name: image_1_prompt
    dtype: string
  - name: image_1
    dtype: image
  - name: image_2_prompt
    dtype: string
  - name: image_2
    dtype: image
  - name: image_3_prompt
    dtype: string
  - name: image_3
    dtype: image
  - name: image_4_prompt
    dtype: string
  - name: image_4
    dtype: image
  - name: image_5_prompt
    dtype: string
  - name: image_5
    dtype: image
  - name: reasoning
    dtype: string
  - name: output
    dtype: string
  splits:
  - name: train
    num_bytes: 1946576155.5
    num_examples: 3996
  - name: test
    num_bytes: 227335843.0
    num_examples: 444
  download_size: 2158171387
  dataset_size: 2173911998.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---