metadata
dataset_info:
features:
- name: image
dtype: image
- name: prompts
sequence: 'null'
- name: messages
sequence: 'null'
splits:
- name: train
num_bytes: 1363415
num_examples: 10
download_size: 1365082
dataset_size: 1363415
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
tags:
- vqasynth
- remyx