|
--- |
|
license: cc-by-4.0 |
|
base_model: NazaGara/NER-fine-tuned-BETO |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- conll2002 |
|
metrics: |
|
- precision |
|
- recall |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: beto-finetuned-ner |
|
results: |
|
- task: |
|
name: Token Classification |
|
type: token-classification |
|
dataset: |
|
name: conll2002 |
|
type: conll2002 |
|
config: es |
|
split: validation |
|
args: es |
|
metrics: |
|
- name: Precision |
|
type: precision |
|
value: 0.8402527075812274 |
|
- name: Recall |
|
type: recall |
|
value: 0.8556985294117647 |
|
- name: F1 |
|
type: f1 |
|
value: 0.8479052823315117 |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.9701834862385321 |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# beto-finetuned-ner |
|
|
|
This model is a fine-tuned version of [NazaGara/NER-fine-tuned-BETO](https://huggingface.co/NazaGara/NER-fine-tuned-BETO) on the conll2002 dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.2248 |
|
- Precision: 0.8403 |
|
- Recall: 0.8557 |
|
- F1: 0.8479 |
|
- Accuracy: 0.9702 |
|
|
|
## Model description |
|
|
|
Este modelo está basado en BETO, que es un modelo de lenguaje preentrenado para el español similar a BERT. BETO fue entrenado inicialmente en grandes cantidades de texto en español. |
|
Posteriormente, este modelo toma la arquitectura y pesos preentrenados de BETO y los ajusta aún más en la tarea específica de Reconocimiento de Entidades Nombradas (NER) utilizando el conjunto de datos conll2002. |
|
Este modelo ajustado puede usarse para anotar automáticamente nuevos textos en español, asignando etiquetas de entidad nombradas. |
|
|
|
## How to Use |
|
|
|
```python |
|
from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("JoshuaAAX/beto-finetuned-ner") |
|
model = AutoModelForTokenClassification.from_pretrained("JoshuaAAX/beto-finetuned-ner") |
|
|
|
|
|
text = "La Federación nacional de cafeteros de Colombia es una entidad del estado. El primer presidente el Dr Augusto Guerra contó con el aval de la Asociación Colombiana de Aviación." |
|
|
|
|
|
ner_pipeline= pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="max") |
|
ner_pipeline(text) |
|
``` |
|
|
|
## Training data |
|
|
|
| Abbreviation | Description | |
|
|:-------------:|:-------------:| |
|
| O | Outside of NE | |
|
| PER | Person’s name | |
|
| ORG | Organization | |
|
| LOC | Location | |
|
| MISC | Miscellaneous | |
|
|
|
|
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 2e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 16 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
| 0.0512 | 1.0 | 521 | 0.1314 | 0.8328 | 0.8562 | 0.8443 | 0.9703 | |
|
| 0.0305 | 2.0 | 1042 | 0.1549 | 0.8318 | 0.8442 | 0.8380 | 0.9688 | |
|
| 0.0193 | 3.0 | 1563 | 0.1498 | 0.8513 | 0.8578 | 0.8545 | 0.9708 | |
|
| 0.0148 | 4.0 | 2084 | 0.1810 | 0.8363 | 0.8442 | 0.8403 | 0.9682 | |
|
| 0.0112 | 5.0 | 2605 | 0.1904 | 0.8412 | 0.8529 | 0.8470 | 0.9703 | |
|
| 0.0078 | 6.0 | 3126 | 0.1831 | 0.8364 | 0.8539 | 0.8450 | 0.9708 | |
|
| 0.0058 | 7.0 | 3647 | 0.2060 | 0.8419 | 0.8543 | 0.8481 | 0.9701 | |
|
| 0.0049 | 8.0 | 4168 | 0.2111 | 0.8357 | 0.8541 | 0.8448 | 0.9697 | |
|
| 0.0037 | 9.0 | 4689 | 0.2255 | 0.8371 | 0.8504 | 0.8437 | 0.9692 | |
|
| 0.0031 | 10.0 | 5210 | 0.2248 | 0.8403 | 0.8557 | 0.8479 | 0.9702 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.0 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|