com.johnsnowlabs.nlp.embeddings.LongformerEmbeddings.scala Maven / Gradle / Ivy
The newest version!
/*
* 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.ai.RoBerta
import com.johnsnowlabs.ml.onnx.{OnnxWrapper, ReadOnnxModel, WriteOnnxModel}
import com.johnsnowlabs.ml.tensorflow._
import com.johnsnowlabs.ml.util.LoadExternalModel.{
loadTextAsset,
modelSanityCheck,
notSupportedEngineError
}
import com.johnsnowlabs.ml.util.{ModelArch, ONNX, TensorFlow}
import com.johnsnowlabs.nlp._
import com.johnsnowlabs.nlp.annotators.common._
import com.johnsnowlabs.nlp.annotators.tokenizer.bpe.BpeTokenizer
import com.johnsnowlabs.nlp.serialization.MapFeature
import com.johnsnowlabs.storage.HasStorageRef
import org.apache.spark.broadcast.Broadcast
import org.apache.spark.ml.param.{IntArrayParam, IntParam}
import org.apache.spark.ml.util.Identifiable
import org.apache.spark.sql.{DataFrame, SparkSession}
/** Longformer is a transformer model for long documents. The Longformer model was presented in
* [[https://arxiv.org/pdf/2004.05150.pdf Longformer: The Long-Document Transformer]] by Iz
* Beltagy, Matthew E. Peters, Arman Cohan. longformer-base-4096 is a BERT-like model started
* from the RoBERTa checkpoint and pretrained for MLM on long documents. It supports sequences of
* length up to 4,096.
*
* Pretrained models can be loaded with `pretrained` of the companion object:
* {{{
* val embeddings = LongformerEmbeddings.pretrained()
* .setInputCols("document", "token")
* .setOutputCol("embeddings")
* }}}
* The default model is `"longformer_base_4096"`, if no name is provided. For available
* pretrained models please see the [[https://sparknlp.org/models?task=Embeddings Models Hub]].
*
* For some examples of usage, see
* [[https://github.com/JohnSnowLabs/spark-nlp/tree/master/src/test/scala/com/johnsnowlabs/nlp/embeddings/LongformerEmbeddingsTestSpec.scala LongformerEmbeddingsTestSpec]].
* To see which models are compatible and how to import them see
* [[https://github.com/JohnSnowLabs/spark-nlp/discussions/5669]].
*
* '''Paper Abstract:'''
*
* ''Transformer-based models are unable to process long sequences due to their self-attention
* operation, which scales quadratically with the sequence length. To address this limitation, we
* introduce the Longformer with an attention mechanism that scales linearly with sequence
* length, making it easy to process documents of thousands of tokens or longer. Longformer's
* attention mechanism is a drop-in replacement for the standard self-attention and combines a
* local windowed attention with a task motivated global attention. Following prior work on
* long-sequence transformers, we evaluate Longformer on character-level language modeling and
* achieve state-of-the-art results on text8 and enwik8. In contrast to most prior work, we also
* pretrain Longformer and finetune it on a variety of downstream tasks. Our pretrained
* Longformer consistently outperforms RoBERTa on long document tasks and sets new
* state-of-the-art results on WikiHop and TriviaQA. We finally introduce the
* Longformer-Encoder-Decoder (LED), a Longformer variant for supporting long document generative
* sequence-to-sequence tasks, and demonstrate its effectiveness on the arXiv summarization
* dataset.''
*
* The original code can be found ```here``` [[https://github.com/allenai/longformer]].
*
* ==Example==
* {{{
* import spark.implicits._
* import com.johnsnowlabs.nlp.base._
* import com.johnsnowlabs.nlp.annotator._
* import org.apache.spark.ml.Pipeline
*
* val documentAssembler = new DocumentAssembler()
* .setInputCol("text")
* .setOutputCol("document")
*
* val tokenizer = new Tokenizer()
* .setInputCols(Array("document"))
* .setOutputCol("token")
*
* val embeddings = LongformerEmbeddings.pretrained()
* .setInputCols("document", "token")
* .setOutputCol("embeddings")
* .setCaseSensitive(true)
*
* val embeddingsFinisher = new EmbeddingsFinisher()
* .setInputCols("embeddings")
* .setOutputCols("finished_embeddings")
* .setOutputAsVector(true)
* .setCleanAnnotations(false)
*
* val pipeline = new Pipeline()
* .setStages(Array(
* documentAssembler,
* tokenizer,
* 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|
* +--------------------------------------------------------------------------------+
* |[0.18792399764060974,-0.14591649174690247,0.20547787845134735,0.1468472778797...|
* |[0.22845706343650818,0.18073144555091858,0.09725798666477203,-0.0417917296290...|
* |[0.07037967443466187,-0.14801117777824402,-0.03603338822722435,-0.17893412709...|
* |[-0.08734266459941864,0.2486150562763214,-0.009067727252840996,-0.24408400058...|
* |[0.22409197688102722,-0.4312366545200348,0.1401449590921402,0.356410235166549...|
* +--------------------------------------------------------------------------------+
* }}}
*
* @see
* [[com.johnsnowlabs.nlp.annotators.classifier.dl.LongformerForTokenClassification LongformerForTokenClassification]]
* for Longformer embeddings with a token classification layer on top
* @see
* [[https://sparknlp.org/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 LongformerEmbeddings(override val uid: String)
extends AnnotatorModel[LongformerEmbeddings]
with HasBatchedAnnotate[LongformerEmbeddings]
with WriteTensorflowModel
with HasEmbeddingsProperties
with HasStorageRef
with HasCaseSensitiveProperties
with HasEngine {
/** Annotator reference id. Used to identify elements in metadata or to refer to this annotator
* type
*/
def this() = this(Identifiable.randomUID("LONGFORMER_EMBEDDINGS"))
def sentenceStartTokenId: Int = {
$$(vocabulary)("")
}
def sentenceEndTokenId: Int = {
$$(vocabulary)("")
}
def padTokenId: Int = {
$$(vocabulary)("")
}
/** Vocabulary used to encode the words to ids with bpeTokenizer.encode
*
* @group param
*/
val vocabulary: MapFeature[String, Int] = new MapFeature(this, "vocabulary").setProtected()
/** @group setParam */
def setVocabulary(value: Map[String, Int]): this.type = set(vocabulary, value)
/** Holding merges.txt coming from Longformer model
*
* @group param
*/
val merges: MapFeature[(String, String), Int] = new MapFeature(this, "merges").setProtected()
/** @group setParam */
def setMerges(value: Map[(String, String), Int]): this.type = set(merges, value)
/** 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()")
/** @group setParam */
def setConfigProtoBytes(bytes: Array[Int]): LongformerEmbeddings.this.type =
set(this.configProtoBytes, bytes)
/** @group getParam */
def getConfigProtoBytes: Option[Array[Byte]] = get(this.configProtoBytes).map(_.map(_.toByte))
/** Max sentence length to process (Default: `128`)
*
* @group param
*/
val maxSentenceLength =
new IntParam(this, "maxSentenceLength", "Max sentence length to process")
/** @group setParam */
def setMaxSentenceLength(value: Int): this.type = {
require(
value <= 4096,
"Longformer models do not support sequences longer than 4096 because of trainable positional embeddings.")
require(value >= 1, "The maxSentenceLength must be at least 1")
set(maxSentenceLength, value)
this
}
/** @group getParam */
def getMaxSentenceLength: Int = $(maxSentenceLength)
/** It contains TF model signatures for the laded saved model
*
* @group param
*/
val signatures =
new MapFeature[String, String](model = this, name = "signatures").setProtected()
/** @group setParam */
def setSignatures(value: Map[String, String]): this.type = {
set(signatures, value)
this
}
/** @group getParam */
def getSignatures: Option[Map[String, String]] = get(this.signatures)
private var _model: Option[Broadcast[RoBerta]] = None
/** @group setParam */
def setModelIfNotSet(
spark: SparkSession,
tensorflowWrapper: Option[TensorflowWrapper],
onnxWrapper: Option[OnnxWrapper]): LongformerEmbeddings = {
if (_model.isEmpty) {
_model = Some(
spark.sparkContext.broadcast(
new RoBerta(
tensorflowWrapper,
onnxWrapper,
None,
sentenceStartTokenId,
sentenceEndTokenId,
padTokenId,
configProtoBytes = getConfigProtoBytes,
signatures = getSignatures,
modelArch = ModelArch.wordEmbeddings)))
}
this
}
/** @group getParam */
def getModelIfNotSet: RoBerta = _model.get.value
/** Set Embeddings dimensions for the RoBERTa model. Only possible to set this when the first
* time is saved dimension is not changeable, it comes from RoBERTa config file.
*
* @group setParam
*/
override def setDimension(value: Int): this.type = {
set(this.dimension, value)
}
/** Whether to lowercase tokens or not
*
* @group setParam
*/
override def setCaseSensitive(value: Boolean): this.type = {
set(this.caseSensitive, value)
}
setDefault(dimension -> 768, batchSize -> 4, maxSentenceLength -> 1024, caseSensitive -> true)
def tokenizeWithAlignment(tokens: Seq[TokenizedSentence]): Seq[WordpieceTokenizedSentence] = {
val bpeTokenizer =
BpeTokenizer.forModel("roberta", merges = $$(merges), vocab = $$(vocabulary))
tokens.map { tokenIndex =>
// filter empty and only whitespace tokens
val bertTokens =
tokenIndex.indexedTokens.filter(x => x.token.nonEmpty && !x.token.equals(" ")).map {
token =>
val content = if ($(caseSensitive)) token.token else token.token.toLowerCase()
val sentenceBegin = token.begin
val sentenceEnd = token.end
val sentenceIndex = tokenIndex.sentenceIndex
val result =
bpeTokenizer.tokenize(Sentence(content, sentenceBegin, sentenceEnd, sentenceIndex))
if (result.nonEmpty) result.head else IndexedToken("")
}
val wordpieceTokens =
bertTokens.flatMap(token => bpeTokenizer.encode(token)).take($(maxSentenceLength))
WordpieceTokenizedSentence(wordpieceTokens)
}
}
/** 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) =>
TokenizedWithSentence.unpack(annotations).toArray.map(x => (x, i))
}
// Tokenize sentences
val tokenizedSentences = tokenizeWithAlignment(sentencesWithRow.map(_._1))
// Process all sentences
val sentenceWordEmbeddings = getModelIfNotSet.predict(
tokenizedSentences,
sentencesWithRow.map(_._1),
$(batchSize),
$(maxSentenceLength),
$(caseSensitive))
// Group resulting annotations by rows. If there are not sentences in a given row, return empty sequence
batchedAnnotations.indices.map(rowIndex => {
val rowEmbeddings = sentenceWordEmbeddings
// 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 (rowEmbeddings.nonEmpty)
WordpieceEmbeddingsSentence.pack(rowEmbeddings)
else
Seq.empty[Annotation]
})
}
override protected def afterAnnotate(dataset: DataFrame): DataFrame = {
dataset.withColumn(
getOutputCol,
wrapEmbeddingsMetadata(dataset.col(getOutputCol), $(dimension), Some($(storageRef))))
}
/** Input Annotator Types: DOCUMENT, TOKEN
*
* @group anno
*/
override val inputAnnotatorTypes: Array[String] =
Array(AnnotatorType.DOCUMENT, AnnotatorType.TOKEN)
/** Output Annotator Types: WORD_EMBEDDINGS
*
* @group anno
*/
override val outputAnnotatorType: AnnotatorType = AnnotatorType.WORD_EMBEDDINGS
override def onWrite(path: String, spark: SparkSession): Unit = {
super.onWrite(path, spark)
writeTensorflowModelV2(
path,
spark,
getModelIfNotSet.tensorflowWrapper.get,
"_longformer",
LongformerEmbeddings.tfFile,
configProtoBytes = getConfigProtoBytes)
}
}
trait ReadablePretrainedLongformerModel
extends ParamsAndFeaturesReadable[LongformerEmbeddings]
with HasPretrained[LongformerEmbeddings] {
override val defaultModelName: Some[String] = Some("longformer_base_4096")
/** Java compliant-overrides */
override def pretrained(): LongformerEmbeddings = super.pretrained()
override def pretrained(name: String): LongformerEmbeddings = super.pretrained(name)
override def pretrained(name: String, lang: String): LongformerEmbeddings =
super.pretrained(name, lang)
override def pretrained(name: String, lang: String, remoteLoc: String): LongformerEmbeddings =
super.pretrained(name, lang, remoteLoc)
}
trait ReadLongformerDLModel extends ReadTensorflowModel {
this: ParamsAndFeaturesReadable[LongformerEmbeddings] =>
override val tfFile: String = "longformer_tensorflow"
def readModel(instance: LongformerEmbeddings, path: String, spark: SparkSession): Unit = {
val tf = readTensorflowModel(path, spark, "_longformer_tf", initAllTables = false)
instance.setModelIfNotSet(spark, Some(tf), None)
}
addReader(readModel)
def loadSavedModel(modelPath: String, spark: SparkSession): LongformerEmbeddings = {
val (localModelPath, detectedEngine) = modelSanityCheck(modelPath)
val vocabs = loadTextAsset(localModelPath, "vocab.txt").zipWithIndex.toMap
val bytePairs = loadTextAsset(localModelPath, "merges.txt")
.map(_.split(" "))
.filter(w => w.length == 2)
.map { case Array(c1, c2) => (c1, c2) }
.zipWithIndex
.toMap
/*Universal parameters for all engines*/
val annotatorModel = new LongformerEmbeddings()
.setVocabulary(vocabs)
.setMerges(bytePairs)
annotatorModel.set(annotatorModel.engine, detectedEngine)
detectedEngine match {
case TensorFlow.name =>
val (wrapper, signatures) =
TensorflowWrapper.read(localModelPath, zipped = false, useBundle = true)
val _signatures = signatures match {
case Some(s) => s
case None => throw new Exception("Cannot load signature definitions from model!")
}
/** the order of setSignatures is important if we use getSignatures inside
* setModelIfNotSet
*/
annotatorModel
.setSignatures(_signatures)
.setModelIfNotSet(spark, Some(wrapper), None)
case _ =>
throw new Exception(notSupportedEngineError)
}
annotatorModel
}
}
/** This is the companion object of [[LongformerEmbeddings]]. Please refer to that class for the
* documentation.
*/
object LongformerEmbeddings extends ReadablePretrainedLongformerModel with ReadLongformerDLModel