Arabic DistilBERT Embeddings (from Geotrend)


Pretrained DistilBERT Embeddings model, uploaded to Hugging Face, adapted and imported into Spark NLP. distilbert-base-ar-cased is a Arabic model orginally trained by Geotrend.


How to use

documentAssembler = DocumentAssembler() \
.setInputCol("text") \

tokenizer = Tokenizer() \
.setInputCols("document") \

embeddings = DistilBertEmbeddings.pretrained("distilbert_embeddings_distilbert_base_ar_cased","ar") \
.setInputCols(["document", "token"]) \

pipeline = Pipeline(stages=[documentAssembler, tokenizer, embeddings])

data = spark.createDataFrame([["أنا أحب شرارة NLP"]]).toDF("text")

result =
val documentAssembler = new DocumentAssembler() 

val tokenizer = new Tokenizer() 

val embeddings = DistilBertEmbeddings.pretrained("distilbert_embeddings_distilbert_base_ar_cased","ar") 
.setInputCols(Array("document", "token")) 

val pipeline = new Pipeline().setStages(Array(documentAssembler, tokenizer, embeddings))

val data = Seq("أنا أحب شرارة NLP").toDF("text")

val result =
import nlu
nlu.load("ar.embed.distilbert").predict("""أنا أحب شرارة NLP""")

Model Information

Model Name: distilbert_embeddings_distilbert_base_ar_cased
Compatibility: Spark NLP 3.4.2+
License: Open Source
Edition: Official
Input Labels: [sentence, token]
Output Labels: [bert]
Language: ar
Size: 182.2 MB
Case sensitive: true