metadata
dataset_info:
features:
- name: font
dtype: string
- name: image
dtype: image
- name: augmented_image
dtype: image
- name: bounding_boxes
sequence:
- name: bbox
sequence: int64
length: 4
- name: char
dtype: string
- name: sample_text
dtype: string
splits:
- name: train
num_bytes: 15585563245
num_examples: 1529600
download_size: 14689346774
dataset_size: 15585563245
configs:
- config_name: default
data_files:
- split: train
path: data/train-*