Description
XLNet 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.
xlnet_base_token_classifier_conll03 is a fine-tuned XLNet model that is ready to use for Named Entity Recognition and achieves state-of-the-art performance for the NER task. This model has been trained to recognize four types of entities: location (LOC), organizations (ORG), person (PER), and Miscellaneous (MISC).
We used TFXLNetForTokenClassification to train this model and used XlnetForTokenClassification
annotator in Spark NLP 🚀 for prediction at scale!
Predicted Entities
How to use
document_assembler = DocumentAssembler() \
.setInputCol('text') \
.setOutputCol('document')
tokenizer = Tokenizer() \
.setInputCols(['document']) \
.setOutputCol('token')
tokenClassifier = XlnetForTokenClassification \
.pretrained('xlnet_base_token_classifier_conll03', 'en') \
.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([['My name is John!']]).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 = XlnetForTokenClassification.pretrained("xlnet_base_token_classifier_conll03", "en")
.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["My name is John!"].toDS.toDF("text")
val result = pipeline.fit(example).transform(example)
import nlu
nlu.load("en.classify.token_xlnet_base_token_classifier_conll03").predict("""My name is John!""")
Results
+------------------------------------------------------------------------------------+
|result |
+------------------------------------------------------------------------------------+
|[B-PER, I-PER, O, O, O, B-LOC, O, O, O, B-LOC, O, O, O, O, B-PER, O, O, O, O, B-LOC]|
+------------------------------------------------------------------------------------+
Model Information
Model Name: | xlnet_base_token_classifier_conll03 |
Compatibility: | Spark NLP 3.3.0+ |
License: | Open Source |
Edition: | Official |
Input Labels: | [token, document] |
Output Labels: | [ner] |
Language: | en |
Case sensitive: | true |
Max sentense length: | 512 |
Data Source
https://www.clips.uantwerpen.be/conll2003/ner/
Benchmarking
precision recall f1-score support
B-LOC 0.94 0.97 0.96 1837
B-MISC 0.84 0.92 0.88 922
B-ORG 0.89 0.94 0.91 1341
B-PER 0.96 0.97 0.97 1842
I-LOC 0.92 0.93 0.92 257
I-MISC 0.84 0.81 0.82 346
I-ORG 0.93 0.89 0.91 751
I-PER 0.96 0.98 0.97 1307
O 1.00 0.99 1.00 42759
accuracy 0.98 51362
macro avg 0.92 0.93 0.93 51362
weighted avg 0.99 0.98 0.99 51362
processed 51362 tokens with 5942 phrases; found: 6219 phrases; correct: 5550.
accuracy: 94.57%; (non-O)
accuracy: 98.49%; precision: 89.24%; recall: 93.40%; FB1: 91.28
LOC: precision: 93.30%; recall: 96.30%; FB1: 94.78 1896
MISC: precision: 79.47%; recall: 88.18%; FB1: 83.60 1023
ORG: precision: 84.71%; recall: 90.08%; FB1: 87.31 1426
PER: precision: 93.92%; recall: 95.55%; FB1: 94.73 1874