metadata
dataset_info:
features:
- name: id
dtype: int64
- name: image
dtype: string
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 80860880
num_examples: 31970
download_size: 36617797
dataset_size: 80860880
configs:
- config_name: default
data_files:
- split: train
path: data/train-*