metadata
dataset_info:
features:
- name: conversation_id
dtype: string
- name: message_id
dtype: string
- name: user_prompt
dtype: string
- name: assistant_response
dtype: string
- name: images
sequence: image
- name: image_urls
sequence: string
- name: json_files
sequence: string
- name: json_contents
sequence: string
- name: num_images
dtype: int64
splits:
- name: train
num_bytes: 27838079
num_examples: 20
download_size: 27731468
dataset_size: 27838079
configs:
- config_name: default
data_files:
- split: train
path: data/train-*