metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_file
dtype: string
- name: prompt
dtype: string
- name: visual_info_type
dtype: string
- name: ocr_text_type
dtype: string
splits:
- name: train
num_bytes: 3282720
num_examples: 15
download_size: 3260678
dataset_size: 3282720
configs:
- config_name: default
data_files:
- split: train
path: data/train-*