Description
This model is based on google’s Flan-T5 Large, and can generate conditional text. Sequence length is 512 tokens.
Predicted Entities
Live Demo Open in Colab Download Copy S3 URI
How to use
document_assembler = DocumentAssembler()\
.setInputCol("prompt")\
.setOutputCol("document_prompt")
med_text_generator = MedicalTextGenerator.pretrained("text_generator_generic_flan_t5_large", "en", "clinical/models")\
.setInputCols("document_prompt")\
.setOutputCol("answer")\
.setMaxNewTokens(256)\
.setDoSample(True)\
.setTopK(3)\
.setRandomSeed(42)
pipeline = Pipeline(stages=[document_assembler, med_text_generator])
data = spark.createDataFrame([["""Classify the following review as negative or positive:
Not a huge fan of her acting, but the movie was actually quite good!"""]]).toDF("prompt")
pipeline.fit(data).transform(data)
val document_assembler = new DocumentAssembler()
.setInputCol("prompt")
.setOutputCol("document_prompt")
val med_text_generator = MedicalTextGenerator.pretrained("text_generator_generic_flan_t5_large", "en", "clinical/models")
.setInputCols("document_prompt")
.setOutputCol("answer")
.setMaxNewTokens(256)
.setDoSample(true)
.setTopK(3)
.setRandomSeed(42)
val pipeline = new Pipeline().setStages(Array(document_assembler, med_text_generator))
val data = Seq(Array("""Classify the following review as negative or positive:
Not a huge fan of her acting, but the movie was actually quite good!""")).toDS.toDF("prompt")
val result = pipeline.fit(data).transform(data)
Results
positive
Model Information
Model Name: | text_generator_generic_flan_t5_large |
Compatibility: | Healthcare NLP 4.3.2+ |
License: | Licensed |
Edition: | Official |
Language: | en |
Size: | 2.9 GB |