metadata
dataset_info:
features:
- name: image
dtype: image
- name: filename
dtype: string
- name: smiles
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: descriptor
dtype: string
splits:
- name: clean
num_bytes: 192571854
num_examples: 30000
- name: large
num_bytes: 468973294
num_examples: 29994
download_size: 963234896
dataset_size: 661545148
configs:
- config_name: default
data_files:
- split: clean
path: data/clean-*
- split: large
path: data/large-*