metadata
dataset_info:
features:
- name: id
dtype: string
- name: image
sequence: string
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 114947
num_examples: 59
- name: validation
num_bytes: 58902
num_examples: 30
download_size: 16719
dataset_size: 173849
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*