All Downloads are FREE. Search and download functionalities are using the official Maven repository.

com.johnsnowlabs.nlp.embeddings.UniversalSentenceEncoder.scala Maven / Gradle / Ivy

/*
 * Copyright 2017-2022 John Snow Labs
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *    http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.johnsnowlabs.nlp.embeddings

import com.johnsnowlabs.ml.tensorflow.{
  ReadTensorflowModel,
  TensorflowUSE,
  TensorflowWrapper,
  WriteTensorflowModel
}
import com.johnsnowlabs.ml.util.LoadExternalModel.{modelSanityCheck, notSupportedEngineError}
import com.johnsnowlabs.ml.util.ModelEngine
import com.johnsnowlabs.nlp.AnnotatorType.{DOCUMENT, SENTENCE_EMBEDDINGS}
import com.johnsnowlabs.nlp._
import com.johnsnowlabs.nlp.annotators.common.SentenceSplit
import com.johnsnowlabs.storage.HasStorageRef
import org.apache.spark.broadcast.Broadcast
import org.apache.spark.ml.param.{BooleanParam, IntArrayParam, IntParam}
import org.apache.spark.ml.util.Identifiable
import org.apache.spark.sql.{DataFrame, SparkSession}

/** The Universal Sentence Encoder encodes text into high dimensional vectors that can be used for
  * text classification, semantic similarity, clustering and other natural language tasks.
  *
  * Pretrained models can be loaded with `pretrained` of the companion object:
  * {{{
  * val useEmbeddings = UniversalSentenceEncoder.pretrained()
  *   .setInputCols("sentence")
  *   .setOutputCol("sentence_embeddings")
  * }}}
  * The default model is `"tfhub_use"`, if no name is provided. For available pretrained models
  * please see the [[https://nlp.johnsnowlabs.com/models?task=Embeddings Models Hub]].
  *
  * For extended examples of usage, see the
  * [[https://github.com/JohnSnowLabs/spark-nlp-workshop/blob/master/tutorials/Certification_Trainings/Public/3.SparkNLP_Pretrained_Models.ipynb Spark NLP Workshop]]
  * and the
  * [[https://github.com/JohnSnowLabs/spark-nlp/blob/master/src/test/scala/com/johnsnowlabs/nlp/embeddings/UniversalSentenceEncoderTestSpec.scala UniversalSentenceEncoderTestSpec]].
  *
  * '''References:'''
  *
  * [[https://arxiv.org/abs/1803.11175 Universal Sentence Encoder]]
  *
  * [[https://tfhub.dev/google/universal-sentence-encoder/2]]
  *
  * '''Paper abstract:'''
  *
  * ''We present models for encoding sentences into embedding vectors that specifically target
  * transfer learning to other NLP tasks. The models are efficient and result in accurate
  * performance on diverse transfer tasks. Two variants of the encoding models allow for
  * trade-offs between accuracy and compute resources. For both variants, we investigate and
  * report the relationship between model complexity, resource consumption, the availability of
  * transfer task training data, and task performance. Comparisons are made with baselines that
  * use word level transfer learning via pretrained word embeddings as well as baselines do not
  * use any transfer learning. We find that transfer learning using sentence embeddings tends to
  * outperform word level transfer. With transfer learning via sentence embeddings, we observe
  * surprisingly good performance with minimal amounts of supervised training data for a transfer
  * task. We obtain encouraging results on Word Embedding Association Tests (WEAT) targeted at
  * detecting model bias. Our pre-trained sentence encoding models are made freely available for
  * download and on TF Hub.''
  *
  * ==Example==
  * {{{
  * import spark.implicits._
  * import com.johnsnowlabs.nlp.base.DocumentAssembler
  * import com.johnsnowlabs.nlp.annotator.SentenceDetector
  * import com.johnsnowlabs.nlp.embeddings.UniversalSentenceEncoder
  * import com.johnsnowlabs.nlp.EmbeddingsFinisher
  * import org.apache.spark.ml.Pipeline
  *
  * val documentAssembler = new DocumentAssembler()
  *   .setInputCol("text")
  *   .setOutputCol("document")
  *
  * val sentence = new SentenceDetector()
  *   .setInputCols("document")
  *   .setOutputCol("sentence")
  *
  * val embeddings = UniversalSentenceEncoder.pretrained()
  *   .setInputCols("sentence")
  *   .setOutputCol("sentence_embeddings")
  *
  * val embeddingsFinisher = new EmbeddingsFinisher()
  *   .setInputCols("sentence_embeddings")
  *   .setOutputCols("finished_embeddings")
  *   .setOutputAsVector(true)
  *   .setCleanAnnotations(false)
  *
  * val pipeline = new Pipeline()
  *   .setStages(Array(
  *     documentAssembler,
  *     sentence,
  *     embeddings,
  *     embeddingsFinisher
  *   ))
  *
  * val data = Seq("This is a sentence.").toDF("text")
  * val result = pipeline.fit(data).transform(data)
  *
  * result.selectExpr("explode(finished_embeddings) as result").show(5, 80)
  * +--------------------------------------------------------------------------------+
  * |                                                                          result|
  * +--------------------------------------------------------------------------------+
  * |[0.04616805538535118,0.022307956591248512,-0.044395286589860916,-0.0016493503...|
  * +--------------------------------------------------------------------------------+
  * }}}
  *
  * @see
  *   [[https://nlp.johnsnowlabs.com/docs/en/annotators Annotators Main Page]] for a list of
  *   transformer based embeddings
  * @groupname anno Annotator types
  * @groupdesc anno
  *   Required input and expected output annotator types
  * @groupname Ungrouped Members
  * @groupname param Parameters
  * @groupname setParam Parameter setters
  * @groupname getParam Parameter getters
  * @groupname Ungrouped Members
  * @groupprio param  1
  * @groupprio anno  2
  * @groupprio Ungrouped 3
  * @groupprio setParam  4
  * @groupprio getParam  5
  * @groupdesc param
  *   A list of (hyper-)parameter keys this annotator can take. Users can set and get the
  *   parameter values through setters and getters, respectively.
  */
class UniversalSentenceEncoder(override val uid: String)
    extends AnnotatorModel[UniversalSentenceEncoder]
    with HasBatchedAnnotate[UniversalSentenceEncoder]
    with HasEmbeddingsProperties
    with HasStorageRef
    with WriteTensorflowModel
    with HasEngine {

  /** Annotator reference id. Used to identify elements in metadata or to refer to this annotator
    * type
    */
  def this() = this(Identifiable.randomUID("UNIVERSAL_SENTENCE_ENCODER"))

  /** Output annotator type : SENTENCE_EMBEDDINGS
    *
    * @group anno
    */
  override val outputAnnotatorType: AnnotatorType = SENTENCE_EMBEDDINGS

  /** Input annotator type : DOCUMENT
    *
    * @group anno
    */
  override val inputAnnotatorTypes: Array[AnnotatorType] = Array(DOCUMENT)

  /** Number of embedding dimensions (Default: `512`)
    *
    * @group param
    */
  override val dimension = new IntParam(this, "dimension", "Number of embedding dimensions")

  /** ConfigProto from tensorflow, serialized into byte array. Get with
    * config_proto.SerializeToString()
    *
    * @group param
    */
  val configProtoBytes = new IntArrayParam(
    this,
    "configProtoBytes",
    "ConfigProto from tensorflow, serialized into byte array. Get with config_proto.SerializeToString()")

  /** Whether to load SentencePiece ops file which is required only by multi-lingual models
    * (Default: `false`). This is not changeable after it's set with a pretrained model nor it is
    * compatible with Windows.
    *
    * @group param
    */
  val loadSP = new BooleanParam(
    this,
    "loadSP",
    "Whether to load SentencePiece ops file which is required only by multi-lingual models. " +
      "This is not changeable after it's set with a pretrained model nor it is compatible with Windows.")

  /** Whether to load SentencePiece ops file which is required only by multi-lingual models.
    *
    * @group setParam
    */
  def setLoadSP(value: Boolean): this.type = {
    if (get(loadSP).isEmpty)
      set(this.loadSP, value)
    this
  }

  /** Whether to load SentencePiece ops file which is required only by multi-lingual models.
    *
    * @group getParam
    */
  def getLoadSP: Boolean = $(loadSP)

  /** ConfigProto from tensorflow, serialized into byte array. Get with
    * config_proto.SerializeToString()
    *
    * @group setParam
    */
  def setConfigProtoBytes(bytes: Array[Int]): UniversalSentenceEncoder.this.type =
    set(this.configProtoBytes, bytes)

  /** ConfigProto from tensorflow, serialized into byte array. Get with
    * config_proto.SerializeToString()
    *
    * @group getParam
    */
  def getConfigProtoBytes: Option[Array[Byte]] =
    get(this.configProtoBytes).map(_.map(_.toByte))

  private var _model: Option[Broadcast[TensorflowUSE]] = None

  /** @group getParam */
  def getModelIfNotSet: TensorflowUSE = _model.get.value

  /** @group setParam */
  def setModelIfNotSet(spark: SparkSession, tensorflow: TensorflowWrapper): this.type = {
    if (_model.isEmpty) {

      _model = Some(
        spark.sparkContext.broadcast(
          new TensorflowUSE(
            tensorflow,
            configProtoBytes = getConfigProtoBytes,
            loadSP = getLoadSP)))
    }
    this
  }

  setDefault(dimension -> 512, storageRef -> "tfhub_use", loadSP -> false, batchSize -> 2)

  /** Takes a document and annotations and produces new annotations of this annotator's annotation
    * type
    *
    * @param batchedAnnotations
    *   Annotations that correspond to inputAnnotationCols generated by previous annotators if any
    * @return
    *   any number of annotations processed for every input annotation. Not necessary one to one
    *   relationship
    */
  override def batchAnnotate(batchedAnnotations: Seq[Array[Annotation]]): Seq[Seq[Annotation]] = {

    // Unpack annotations and zip each sentence to the index or the row it belongs to
    val sentencesWithRow = batchedAnnotations.zipWithIndex
      .flatMap { case (annotations, i) => SentenceSplit.unpack(annotations).map(x => (x, i)) }

    val nonEmptySentences = sentencesWithRow.map(_._1).filter(_.content.nonEmpty)
    val allAnnotations =
      if (nonEmptySentences.nonEmpty)
        getModelIfNotSet.predict(nonEmptySentences, $(batchSize))
      else Seq.empty[Annotation]

    // Group resulting annotations by rows. If there are not sentences in a given row, return empty sequence
    batchedAnnotations.indices.map(rowIndex => {
      val rowAnnotations = allAnnotations
        // zip each annotation with its corresponding row index
        .zip(sentencesWithRow)
        // select the sentences belonging to the current row
        .filter(_._2._2 == rowIndex)
        // leave the annotation only
        .map(_._1)

      if (rowAnnotations.nonEmpty)
        rowAnnotations
      else
        Seq.empty[Annotation]
    })
  }

  override protected def afterAnnotate(dataset: DataFrame): DataFrame = {
    dataset.withColumn(
      getOutputCol,
      wrapSentenceEmbeddingsMetadata(
        dataset.col(getOutputCol),
        $(dimension),
        Some($(storageRef))))
  }

  override def onWrite(path: String, spark: SparkSession): Unit = {
    super.onWrite(path, spark)
    writeTensorflowModelV2(
      path,
      spark,
      getModelIfNotSet.tensorflow,
      "_use",
      UniversalSentenceEncoder.tfFile,
      configProtoBytes = getConfigProtoBytes)
  }

}

trait ReadablePretrainedUSEModel
    extends ParamsAndFeaturesReadable[UniversalSentenceEncoder]
    with HasPretrained[UniversalSentenceEncoder] {
  override val defaultModelName: Some[String] = Some("tfhub_use")

  /** Java compliant-overrides */
  override def pretrained(): UniversalSentenceEncoder = super.pretrained()

  override def pretrained(name: String): UniversalSentenceEncoder = super.pretrained(name)

  override def pretrained(name: String, lang: String): UniversalSentenceEncoder =
    super.pretrained(name, lang)

  override def pretrained(
      name: String,
      lang: String,
      remoteLoc: String): UniversalSentenceEncoder =
    super.pretrained(name, lang, remoteLoc)
}

trait ReadUSEDLModel extends ReadTensorflowModel {
  this: ParamsAndFeaturesReadable[UniversalSentenceEncoder] =>

  /*Needs to point to an actual folder rather than a .pb file*/
  override val tfFile: String = "use_tensorflow"

  def readTensorflow(
      instance: UniversalSentenceEncoder,
      path: String,
      spark: SparkSession): Unit = {
    val loadSP = instance.getLoadSP
    val tf =
      readTensorflowWithSPModel(path, spark, "_use_tf", initAllTables = true, loadSP = loadSP)
    instance.setModelIfNotSet(spark, tf)
  }

  addReader(readTensorflow)

  def loadSavedModel(
      modelPath: String,
      spark: SparkSession,
      loadSP: Boolean = false): UniversalSentenceEncoder = {

    val (localModelPath, detectedEngine) = modelSanityCheck(modelPath)

    /*Universal parameters for all engines*/
    val annotatorModel = new UniversalSentenceEncoder()
      .setLoadSP(loadSP)

    annotatorModel.set(annotatorModel.engine, detectedEngine)

    detectedEngine match {
      case ModelEngine.tensorflow =>
        val wrapper =
          TensorflowWrapper.readWithSP(
            localModelPath,
            zipped = false,
            useBundle = true,
            tags = Array("serve"),
            initAllTables = true,
            loadSP = loadSP)

        /** the order of setSignatures is important if we use getSignatures inside
          * setModelIfNotSet
          */
        annotatorModel
          .setModelIfNotSet(spark, wrapper)

      case _ =>
        throw new Exception(notSupportedEngineError)
    }

    annotatorModel

  }
}

/** This is the companion object of [[UniversalSentenceEncoder]]. Please refer to that class for
  * the documentation.
  */
object UniversalSentenceEncoder extends ReadablePretrainedUSEModel with ReadUSEDLModel




© 2015 - 2024 Weber Informatics LLC | Privacy Policy