README.md

---
language: "pt"
widget:
- text: "Tinha uma pedra no meio do caminho."
- text: "Vamos tomar um café quentinho?"
- text: "Como você se chama?"
datasets: 
- MacMorpho
---

# POS-Tagger Portuguese

We fine-tuned the [BERTimbau](https://github.com/neuralmind-ai/portuguese-bert/) model with the MacMorpho corpus for the Post-Tagger task, with 10 epochs, achieving a general F1-Score of 0.9837.

Metrics:

```
              Precision  Recall  F1    Suport
accuracy                         0.98  33729
macro avg     0.96       0.96    0.96  33729
weighted avg  0.98       0.98    0.98  33729

F1:  0.9837050718155329 Accuracy:  0.98372320555012
```

## Questions?

Please, post a Github issue on the [NLP Portuguese POS-Tagger](https://github.com/lisaterumi/nlp-portuguese-postagger).