chore: improve docs

README.md

@@ -1,3 +1,47 @@
----
-license: mit
----
+# 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}
+}
+```