Description
Pretrained RoBERTa Embeddings model, uploaded to Hugging Face, adapted and imported into Spark NLP. am-roberta
is a Amharic model orginally trained by uhhlt
.
How to use
documentAssembler = DocumentAssembler() \
.setInputCol("text") \
.setOutputCol("document")
tokenizer = Tokenizer() \
.setInputCols("document") \
.setOutputCol("token")
embeddings = RoBertaEmbeddings.pretrained("roberta_embeddings_am_roberta","am") \
.setInputCols(["document", "token"]) \
.setOutputCol("embeddings")
pipeline = Pipeline(stages=[documentAssembler, tokenizer, embeddings])
data = spark.createDataFrame([["ስካርቻ nlp እወዳለሁ"]]).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 = RoBertaEmbeddings.pretrained("roberta_embeddings_am_roberta","am")
.setInputCols(Array("document", "token"))
.setOutputCol("embeddings")
val pipeline = new Pipeline().setStages(Array(documentAssembler, tokenizer, embeddings))
val data = Seq("ስካርቻ nlp እወዳለሁ").toDF("text")
val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("am.embed.am_roberta").predict("""ስካርቻ nlp እወዳለሁ""")
Model Information
Model Name: | roberta_embeddings_am_roberta |
Compatibility: | Spark NLP 3.4.2+ |
License: | Open Source |
Edition: | Official |
Input Labels: | [sentence, token] |
Output Labels: | [bert] |
Language: | am |
Size: | 1.6 GB |
Case sensitive: | true |
References
- https://huggingface.co/uhhlt/am-roberta
- https://github.com/uhh-lt/amharicmodels
- https://github.com/uhh-lt/amharicmodels
- https://www.mdpi.com/1999-5903/13/11/275