---
license: mit
language:
- en
- fr
- de
- la
- it
base_model:
- sven-nm/XLM-R-for-classics
pipeline_tag: token-classification
---

This model is fine-tuned for lemma recognition on commentaries texts for 40 epochs, beating all our other models. 
Lemma recognition consists in recognitizing the _lemma_ in a gloss, that is the part of the gloss which contains the words being commented upon. 
This model is intended to detect both lemmas and verse numbers.
It can be used in a normal token classification pipeline.