metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: window_size
dtype: int32
- name: window_index
dtype: int32
- name: line_ids
dtype: string
- name: line_reading_orders
dtype: string
- name: region_id
dtype: string
- name: region_reading_order
dtype: int32
- name: region_type
dtype: string
- name: filename
dtype: string
- name: project
dtype: string
splits:
- name: train
num_bytes: 295772707
num_examples: 725
download_size: 295703362
dataset_size: 295772707
configs:
- config_name: default
data_files:
- split: train
path: data/train-*