BERT Token Classification - BETO Spanish Language Understanding (bert_token_classifier_spanish_ner)

Description

BERT Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks.

This model is a fine-tuned on NER-C version of the Spanish BERT cased (BETO) for NER downstream task.

Predicted Entities

  • B-LOC
  • B-MISC
  • B-ORG
  • B-PER
  • I-LOC
  • I-MISC
  • I-ORG
  • I-PER
  • O

Download Copy S3 URI

How to use

document_assembler = DocumentAssembler() \
.setInputCol('text') \
.setOutputCol('document')

tokenizer = Tokenizer() \
.setInputCols(['document']) \
.setOutputCol('token')

tokenClassifier = BertForTokenClassification \
.pretrained('bert_token_classifier_spanish_ner', 'es') \
.setInputCols(['token', 'document']) \
.setOutputCol('ner') \
.setCaseSensitive(True) \
.setMaxSentenceLength(512)

# since output column is IOB/IOB2 style, NerConverter can extract entities
ner_converter = NerConverter() \
.setInputCols(['document', 'token', 'ner']) \
.setOutputCol('entities')

pipeline = Pipeline(stages=[
document_assembler, 
tokenizer,
tokenClassifier,
ner_converter
])

example = spark.createDataFrame([["Me llamo Wolfgang y vivo en Berlin"]]).toDF("text")
result = pipeline.fit(example).transform(example)
val document_assembler = DocumentAssembler() 
.setInputCol("text") 
.setOutputCol("document")

val tokenizer = Tokenizer() 
.setInputCols("document") 
.setOutputCol("token")

val tokenClassifier = BertForTokenClassification.pretrained("bert_token_classifier_spanish_ner", "es")
.setInputCols("document", "token")
.setOutputCol("ner")
.setCaseSensitive(true)
.setMaxSentenceLength(512)

// since output column is IOB/IOB2 style, NerConverter can extract entities
val ner_converter = NerConverter() 
.setInputCols("document", "token", "ner") 
.setOutputCol("entities")

val pipeline = new Pipeline().setStages(Array(document_assembler, tokenizer, tokenClassifier, ner_converter))

val example = Seq.empty["Me llamo Wolfgang y vivo en Berlin"].toDS.toDF("text")

val result = pipeline.fit(example).transform(example)
import nlu
nlu.load("es.classify.token_bert.spanish_ner").predict("""Me llamo Wolfgang y vivo en Berlin""")

Model Information

Model Name: bert_token_classifier_spanish_ner
Compatibility: Spark NLP 3.2.0+
License: Open Source
Edition: Official
Input Labels: [token, document]
Output Labels: [ner]
Language: es
Case sensitive: false
Max sentense length: 512

Data Source

https://huggingface.co/mrm8488/bert-spanish-cased-finetuned-ner

Benchmarking

|                                                      Metric                                                       |  # score  |
| :------------------------------------------------------------------------------------: | :-------: |
| F1                                       | **90.17**  
| Precision                                | **89.86** | 
| Recall                                   | **90.47** |    

## Comparison:

|                                                      Model                                                       |  # F1 score  |Size(MB)|
| :--------------------------------------------------------------------------------------------------------------: | :-------: |:------|
|                                        bert-base-spanish-wwm-cased (BETO)                                        |   88.43   | 421
| [bert-spanish-cased-finetuned-ner (this one)](https://huggingface.co/mrm8488/bert-spanish-cased-finetuned-ner) | **90.17** | 420 |
|                                              Best Multilingual BERT                                              |   87.38   | 681 |
|[TinyBERT-spanish-uncased-finetuned-ner](https://huggingface.co/mrm8488/TinyBERT-spanish-uncased-finetuned-ner) | 70.00 | **55** |