Chinese Deberta Embeddings Cased model (from IDEA-CCNL)

Description

Pretrained DebertaV2ForMaskedLM model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP. Erlangshen-DeBERTa-v2-186M-Chinese-SentencePiece is a Chinese model originally trained by IDEA-CCNL.

Download Copy S3 URI

How to use

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

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

embeddings = DeBertaEmbeddings.pretrained("deberta_embeddings_erlangshen_v2_chinese_sentencepiece","zh") \
    .setInputCols(["document", "token"]) \
    .setOutputCol("embeddings") \
    .setCaseSensitive(True)

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

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 embeddings = DeBertaEmbeddings.pretrained("deberta_embeddings_erlangshen_v2_chinese_sentencepiece","zh")
    .setInputCols(Array("document", "token"))
    .setOutputCol("embeddings")
    .setCaseSensitive(True)

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

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

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

Model Information

Model Name: deberta_embeddings_erlangshen_v2_chinese_sentencepiece
Compatibility: Spark NLP 4.3.0+
License: Open Source
Edition: Official
Input Labels: [sentence, token]
Output Labels: [embeddings]
Language: zh
Size: 445.8 MB
Case sensitive: false

References

https://huggingface.co/IDEA-CCNL/Erlangshen-DeBERTa-v2-186M-Chinese-SentencePiece