metadata
dataset_info:
features:
- name: conversation
dtype: string
- name: image
dtype: string
splits:
- name: train
num_bytes: 2022214849
num_examples: 3167
download_size: 1145047461
dataset_size: 2022214849
configs:
- config_name: default
data_files:
- split: train
path: data/train-*