metadata
dataset_info:
features:
- name: answer
dtype: string
- name: cluster
dtype: int32
- name: image
dtype: image
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 965568274.42
num_examples: 5820
- name: test
num_bytes: 358795481.815
num_examples: 2495
download_size: 1239772291
dataset_size: 1324363756.235
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*