Description
Pretrained BertForQuestionAnswering model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP.bert_large_uncased_whole_word_masking_squad2_p70 is a English model originally trained by pminha.
How to use
documentAssembler = MultiDocumentAssembler() \
.setInputCol(["question", "context"]) \
.setOutputCol(["document_question", "document_context"])
spanClassifier = BertForQuestionAnswering.pretrained("bert_large_uncased_whole_word_masking_squad2_p70","en") \
.setInputCols(["document_question","document_context"]) \
.setOutputCol("answer")
pipeline = Pipeline().setStages([documentAssembler, spanClassifier])
data = spark.createDataFrame([["What framework do I use?","I use spark-nlp."]]).toDF("document_question", "document_context")
pipelineModel = pipeline.fit(data)
pipelineDF = pipelineModel.transform(data)
val documentAssembler = new MultiDocumentAssembler()
.setInputCol(Array("question", "context"))
.setOutputCol(Array("document_question", "document_context"))
val spanClassifier = BertForQuestionAnswering.pretrained("bert_large_uncased_whole_word_masking_squad2_p70", "en")
.setInputCols(Array("document_question","document_context"))
.setOutputCol("answer")
val pipeline = new Pipeline().setStages(Array(documentAssembler, spanClassifier))
val data = Seq("What framework do I use?","I use spark-nlp.").toDS.toDF("document_question", "document_context")
val pipelineModel = pipeline.fit(data)
val pipelineDF = pipelineModel.transform(data)
Model Information
| Model Name: | bert_large_uncased_whole_word_masking_squad2_p70 |
| Compatibility: | Spark NLP 5.5.1+ |
| License: | Open Source |
| Edition: | Official |
| Input Labels: | [document_question, document_context] |
| Output Labels: | [answer] |
| Language: | en |
| Size: | 596.1 MB |
References
https://huggingface.co/pminha/bert-large-uncased-whole-word-masking-squad2-p70