metadata
dataset_info:
features:
- name: image
dtype: string
- name: id
dtype: string
- name: conversations
dtype: string
splits:
- name: train
num_bytes: 4309970
num_examples: 18845
- name: test
num_bytes: 626175
num_examples: 2790
download_size: 924604
dataset_size: 4936145
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*