---
dataset_info:
  features:
  - name: im
    dtype: image
  - name: roman
    dtype: string
  - name: manchu
    dtype: string
  splits:
  - name: train
    num_bytes: 746967.9223744292
    num_examples: 175
  - name: validation
    num_bytes: 187809.0776255708
    num_examples: 44
  download_size: 946637
  dataset_size: 934777.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---