French ALBERT Embeddings (from qwant)

Description

Pretrained ALBERT Embeddings model, uploaded to Hugging Face, adapted and imported into Spark NLP. fralbert-base is a French model orginally trained by qwant.

Download

How to use

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

tokenizer = Tokenizer() \
    .setInputCols("document") \
    .setOutputCol("token")
  
embeddings = AlbertEmbeddings.pretrained("albert_embeddings_fralbert_base","fr") \
    .setInputCols(["document", "token"]) \
    .setOutputCol("embeddings")
    
pipeline = Pipeline(stages=[documentAssembler, tokenizer, embeddings])

data = spark.createDataFrame([["J'adore Spark Nlp"]]).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 = AlbertEmbeddings.pretrained("albert_embeddings_fralbert_base","fr") 
    .setInputCols(Array("document", "token")) 
    .setOutputCol("embeddings")

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

val data = Seq("J'adore Spark Nlp").toDF("text")

val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("fr.embed.albert").predict("""J'adore Spark Nlp""")

Model Information

Model Name: albert_embeddings_fralbert_base
Compatibility: Spark NLP 3.4.2+
License: Open Source
Edition: Official
Input Labels: [sentence, token]
Output Labels: [bert]
Language: fr
Size: 45.9 MB
Case sensitive: false

References

  • https://huggingface.co/qwant/fralbert-base
  • https://arxiv.org/abs/1909.11942
  • https://github.com/google-research/albert
  • https://fr.wikipedia.org/wiki/French_Wikipedia
  • https://hal.archives-ouvertes.fr/hal-03336060