Description
Pretrained BertForSequenceClassification model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP. hatescore-korean-hate-speech is a Korean model originally trained by sgunderscore.
Predicted Entities
종교, None, 여성/가족, 인종/국적, 기타 혐오, 단순 악플, 성소수자, 지역, 연령, 남성
How to use
documentAssembler = DocumentAssembler() \
.setInputCol("text") \
.setOutputCol("document")
tokenizer = Tokenizer() \
.setInputCols("document") \
.setOutputCol("token")
sequenceClassifier_loaded = BertForSequenceClassification.pretrained("bert_classifier_hatescore_korean_hate_speech","ko") \
.setInputCols(["document", "token"]) \
.setOutputCol("class")
pipeline = Pipeline(stages=[documentAssembler, tokenizer,sequenceClassifier_loaded])
data = spark.createDataFrame([["나는 Spark NLP를 좋아합니다"]]).toDF("text")
result = pipeline.fit(data).transform(data)
val documentAssembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val tokenizer = new Tokenizer()
.setInputCols(Array("document"))
.setOutputCol("token")
val sequenceClassifier_loaded = BertForSequenceClassification.pretrained("bert_classifier_hatescore_korean_hate_speech","ko")
.setInputCols(Array("document", "token"))
.setOutputCol("class")
val pipeline = new Pipeline().setStages(Array(documentAssembler, tokenizer,sequenceClassifier_loaded))
val data = Seq("나는 Spark NLP를 좋아합니다").toDF("text")
val result = pipeline.fit(data).transform(data)
Model Information
| Model Name: | bert_classifier_hatescore_korean_hate_speech |
| Compatibility: | Spark NLP 5.1.4+ |
| License: | Open Source |
| Edition: | Official |
| Input Labels: | [document, token] |
| Output Labels: | [class] |
| Language: | ko |
| Size: | 408.5 MB |
| Case sensitive: | true |
| Max sentence length: | 256 |
References
References
- https://huggingface.co/sgunderscore/hatescore-korean-hate-speech