metadata
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
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-*