metadata
dataset_info:
features:
- name: image
dtype: image
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 8806713
num_examples: 30
download_size: 7373646
dataset_size: 8806713
configs:
- config_name: default
data_files:
- split: train
path: data/train-*