---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    sequence: string
  - name: kanji
    dtype: string
  - name: image_path
    dtype: string
  splits:
  - name: train
    num_bytes: 20176738.08
    num_examples: 2860
  - name: test
    num_bytes: 4079955.0
    num_examples: 574
  download_size: 24402583
  dataset_size: 24256693.08
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---