Edit on GitHub
  • John Snow Labs
  • Dec 02, 2022

Spanish BertForMaskedLM Base Cased model (from dccuchile)

  • es
  • open_source
  • bert_embeddings
  • bertformaskedlm

Description

Pretrained BertForMaskedLM model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP. bert-base-spanish-wwm-cased is a Spanish model originally trained by dccuchile.

Download Copy S3 URI

How to use

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

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

bert_loaded = BertEmbeddings.pretrained("bert_embeddings_base_spanish_wwm_cased","es") \
    .setInputCols(["document", "token"]) \
    .setOutputCol("embeddings") \
    .setCaseSensitive(True)

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

data = spark.createDataFrame([["I love Spark NLP"]]).toDF("text")

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

val documentAssembler = new DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")

val tokenizer = new Tokenizer()
    .setInputCols("document")
    .setOutputCol("token")

val bert_loaded = BertEmbeddings.pretrained("bert_embeddings_base_spanish_wwm_cased","es")
    .setInputCols(Array("document", "token"))
    .setOutputCol("embeddings")
    .setCaseSensitive(True)

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

val data = Seq("I love Spark NLP").toDS.toDF("text")

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

import nlu
nlu.load("es.embed.bert.cased_base.by_dccuchile").predict("""I love Spark NLP""")

Model Information

Model Name: bert_embeddings_base_spanish_wwm_cased
Compatibility: Spark NLP 4.2.4+
License: Open Source
Edition: Official
Input Labels: [sentence, token]
Output Labels: [bert]
Language: es
Size: 412.2 MB
Case sensitive: true

References

  • https://huggingface.co/dccuchile/bert-base-spanish-wwm-cased
  • https://github.com/google-research/bert
  • https://github.com/josecannete/spanish-corpora
  • https://github.com/google-research/bert/blob/master/multilingual.md
  • https://users.dcc.uchile.cl/~jperez/beto/uncased_2M/tensorflow_weights.tar.gz
  • https://users.dcc.uchile.cl/~jperez/beto/uncased_2M/pytorch_weights.tar.gz
  • https://users.dcc.uchile.cl/~jperez/beto/cased_2M/tensorflow_weights.tar.gz
  • https://users.dcc.uchile.cl/~jperez/beto/cased_2M/pytorch_weights.tar.gz
  • https://lindat.mff.cuni.cz/repository/xmlui/handle/11234/1-1827
  • https://www.kaggle.com/nltkdata/conll-corpora
  • https://github.com/gchaperon/beto-benchmarks/blob/master/conll2002/dev_results_beto-cased_conll2002.txt
  • https://github.com/facebookresearch/MLDoc
  • https://github.com/gchaperon/beto-benchmarks/blob/master/MLDoc/dev_results_beto-cased_mldoc.txt
  • https://github.com/gchaperon/beto-benchmarks/blob/master/MLDoc/dev_results_beto-uncased_mldoc.txt
  • https://github.com/google-research-datasets/paws/tree/master/pawsx
  • https://github.com/facebookresearch/XNLI
  • https://colab.research.google.com/drive/1uRwg4UmPgYIqGYY4gW_Nsw9782GFJbPt
  • https://www.adere.so/
  • https://imfd.cl/en/
  • https://www.tensorflow.org/tfrc
  • https://users.dcc.uchile.cl/~jperez/papers/pml4dc2020.pdf
  • https://github.com/google-research/bert/blob/master/multilingual.md
  • https://arxiv.org/pdf/1904.09077.pdf
  • https://arxiv.org/pdf/1906.01502.pdf
  • https://arxiv.org/abs/1812.10464
  • https://arxiv.org/pdf/1901.07291.pdf
  • https://arxiv.org/pdf/1904.02099.pdf
  • https://arxiv.org/pdf/1906.01569.pdf
  • https://arxiv.org/abs/1908.11828