metadata
dataset_info:
features:
- name: image
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 200
num_examples: 2
configs:
- config_name: default
data_files:
- split: train
path: data.jsonl