metadata
dataset_info:
features:
- name: band
dtype: string
- name: character
dtype: string
- name: scene
dtype: string
- name: reference
dtype: string
splits:
- name: train
num_bytes: 15467710
num_examples: 7501
download_size: 3111357
dataset_size: 15467710
configs:
- config_name: default
data_files:
- split: train
path: data/train-*