metadata
dataset_info:
features:
- name: image
dtype: string
- name: id
dtype: string
- name: conversations
dtype: string
splits:
- name: train
num_bytes: 4683526
num_examples: 23785
- name: test
num_bytes: 977129
num_examples: 5015
download_size: 1360403
dataset_size: 5660655
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*