metadata
dataset_info:
features:
- name: image
dtype: string
- name: id
dtype: string
- name: conversations
dtype: string
splits:
- name: train
num_bytes: 2011122
num_examples: 11999
- name: test
num_bytes: 504924
num_examples: 3001
download_size: 794377
dataset_size: 2516046
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*