Marathi DistilBERT Embeddings (from DarshanDeshpande)

Description

Pretrained DistilBERT Embeddings model, uploaded to Hugging Face, adapted and imported into Spark NLP. marathi-distilbert is a Marathi model orginally trained by DarshanDeshpande.

Download

How to use

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

tokenizer = Tokenizer() \
    .setInputCols("document") \
    .setOutputCol("token")
  
embeddings = DistilBertEmbeddings.pretrained("distilbert_embeddings_marathi_distilbert","mr") \
    .setInputCols(["document", "token"]) \
    .setOutputCol("embeddings")
    
pipeline = Pipeline(stages=[documentAssembler, tokenizer, embeddings])

data = spark.createDataFrame([["मला स्पार्क एनएलपी आवडते"]]).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 embeddings = DistilBertEmbeddings.pretrained("distilbert_embeddings_marathi_distilbert","mr") 
    .setInputCols(Array("document", "token")) 
    .setOutputCol("embeddings")

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

val data = Seq("मला स्पार्क एनएलपी आवडते").toDF("text")

val result = pipeline.fit(data).transform(data)

Model Information

Model Name: distilbert_embeddings_marathi_distilbert
Compatibility: Spark NLP 3.4.2+
License: Open Source
Edition: Official
Input Labels: [sentence, token]
Output Labels: [bert]
Language: mr
Size: 247.9 MB
Case sensitive: true

References

  • https://huggingface.co/DarshanDeshpande/marathi-distilbert
  • https://github.com/DarshanDeshpande
  • https://www.linkedin.com/in/darshan-deshpande/
  • https://github.com/Baras64
  • http://​www.linkedin.com/in/harsh-abhi