metadata
dataset_info:
features:
- name: id
dtype: string
- name: image_input
dtype: image
- name: prompt
dtype: string
- name: text_output
dtype: string
- name: image_output
dtype: image
- name: model_name
dtype: string
- name: meta_task_name
dtype: string
- name: sub_task_name
dtype: string
splits:
- name: train
num_bytes: 11168806
num_examples: 6
download_size: 11068896
dataset_size: 11168806
configs:
- config_name: default
data_files:
- split: train
path: data/train-*