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: source_image
dtype: image
- name: edited_image
dtype: image
splits:
- name: train
num_bytes: 867310427.015
num_examples: 1499
download_size: 849257331
dataset_size: 867310427.015
configs:
- config_name: default
data_files:
- split: train
path: data/train-*