metadata
dataset_info:
features:
- name: image
dtype: image
- name: messages
dtype: 'null'
splits:
- name: train
num_bytes: 1363335
num_examples: 10
download_size: 1364618
dataset_size: 1363335
configs:
- config_name: default
data_files:
- split: train
path: data/train-*