---
license: cc-by-nc-sa-4.0
language:
  - cu
tags:
  - old-church-slavonic
  - church-slavonic
  - stanza
  - tokenizer
  - lemmatizer
  - lemmatization
  - historical-nlp
  - universal-dependencies
---

# Old Church Slavonic Tokenizer and Lemmatizer

This repository provides retrained Stanza-compatible tokenizer and lemmatizer model weights for Old Church Slavonic.

## Files

```text
models/
├── new-data/
│   ├── tokenize/cu_proiel_tokenizer.pt
│   └── lemma/cu_proiel_nocharlm_lemmatizer.pt
└── combined/
    ├── tokenize/cu_proiel_tokenizer.pt
    └── lemma/cu_proiel_nocharlm_lemmatizer.pt
```

## Model variants

### New-data model

`models/new-data/` contains the tokenizer and lemmatizer trained on the newly annotated Old Church Slavonic dataset.

### Combined model

`models/combined/` contains the tokenizer and lemmatizer trained on a combined corpus including the newly annotated Old Church Slavonic dataset and UD Old Church Slavonic PROIEL.




## License

The model weights are released under CC BY-NC-SA 4.0.