README.md

# BERT pt-BR Persons

This model is fine-tuned to primarily identify Brazilian names, ignoring street and place names, even if they contain a person's name.

![test](./figures/test.png)

## Basic Usage

```python

from transformers import pipeline, BertForTokenClassification, BertTokenizerFast



model_name = "rafola/BERT-base-pt-BR-person"

model = BertForTokenClassification.from_pretrained(model_name)

tokenizer = BertTokenizerFast.from_pretrained(model_name,use_fast=True)



nlp = pipeline("ner", model=model, tokenizer=tokenizer)



result=nlp("Como já dizia seu Zé Ricardo, A Luiza sempre vai atrás de uma encrenca, mesmo com todo o cuidado de tia Eliana com ela.")



print(result)

```

## Citations

If you use our work, please cite:

```

@misc{rafola2025BERTptBRpersons,

  author = {Rafael Vitor Krueger},

  title = {BERT pt-BR Persons: Fine-tuned model to identify brazilian person names},

  year = {2025},

  url = {https://huggingface.co/rafola/BERT-base-pt-BR-person},

}

```

This model has been treined using as base: [neuralmind/bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased)

```

@inproceedings{souza2020bertimbau,

  author    = {F{\'a}bio Souza and

               Rodrigo Nogueira and

               Roberto Lotufo},

  title     = {{BERT}imbau: pretrained {BERT} models for {B}razilian {P}ortuguese},

  booktitle = {9th Brazilian Conference on Intelligent Systems, {BRACIS}, Rio Grande do Sul, Brazil, October 20-23 (to appear)},

  year      = {2020}

}

```