metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 113667809.83200002
num_examples: 15453
- name: test
num_bytes: 12549739.297
num_examples: 1717
download_size: 75076409
dataset_size: 126217549.12900002
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
uv run bin/generate.py