---
language:
- la
metrics:
- accuracy
library_name: spacy
pipeline_tag: token-classification
---
# NER for Latin

Trained using letters from the [Bullinger collection](http://www.bullinger-digital.ch/), based on mbert.

# How to use
```python
import spacy
nlp = spacy.load('./enhg_pipeline')
doc = nlp('Norimberga in proximum quoddam Ulmensibus oppidulum Leypphaim sese contulit, certa spe recuperandae sedis, e qua nuper est detrusus.')
for ent in doc.ents:
  print(ent.text, ent.label_)

# Output:
# Norimberga GEO
# Ulmensibus GEO
```

# Evaluation
- F-Score: 0.8970679975
- Precision: 0.8860135551,
- Recall: 0.9084017688,