README.md

---
dataset_info:
  features:
  - name: im
    dtype: image
  - name: roman
    dtype: string
  - name: manchu
    dtype: string
  splits:
  - name: train
    num_bytes: 305015678
    num_examples: 80306
  - name: validation
    num_bytes: 67762210
    num_examples: 18359
  - name: test
    num_bytes: 789711
    num_examples: 753
  download_size: 370723474
  dataset_size: 373567599
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---