metadata
dataset_info:
features:
- name: key
dtype: string
- name: romanization
dtype: string
- name: hanzi
dtype: string
- name: script_tags
list: string
- name: printed_pages
dtype: string
- name: pdf_pages
dtype: string
- name: n_pages
dtype: int32
- name: page_index
dtype: int32
- name: image
dtype: image
splits:
- name: train
num_bytes: 195056701
num_examples: 76
download_size: 195062582
dataset_size: 195056701
configs:
- config_name: default
data_files:
- split: train
path: data/train-*