Modelo para Reconhecimento de Entidade Nomeadas em português utilizando o modelo spaCy pt_core_news_lg

Link do trabalho no Kaggle: https://www.kaggle.com/datasets/flaviagg/lenerbr .

Criei um Web App que proporciona a comparação dos modelos sm e lg: https://huggingface.co./spaces/flaviaggp/Streamlit_Lener .

Métricas por entidade

Screenshot

Feature Description
Name pt_lg_pipeline
Version 0.0.0
spaCy >=3.4.4,<3.5.0
Default Pipeline tok2vec, ner
Components tok2vec, ner
Vectors 500000 keys, 500000 unique vectors (300 dimensions)
Sources n/a
License n/a
Author n/a

Label Scheme

View label scheme (6 labels for 1 components)
Component Labels
ner JURISPRUDENCIA, LEGISLACAO, LOCAL, ORGANIZACAO, PESSOA, TEMPO

Accuracy

Type Score
ENTS_F 83.79
ENTS_P 83.98
ENTS_R 83.61
TOK2VEC_LOSS 23620.33
NER_LOSS 127975.46
Downloads last month
2
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train flaviaggp/pt_lg_pipeline

Space using flaviaggp/pt_lg_pipeline 1

Evaluation results