metadata
dataset_info:
features:
- name: image
dtype: image
- name: char
dtype: string
- name: unicode
dtype: string
- name: font
dtype: string
- name: font_type
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 2047258850.3594296
num_examples: 458037
- name: test
num_bytes: 446963.6405703971
num_examples: 100
download_size: 1854441785
dataset_size: 2047705814
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*