spark-nlp 5.3.2__py2.py3-none-any.whl → 5.4.0__py2.py3-none-any.whl

sparknlp/annotator/classifier_dl/mpnet_for_token_classification.py

@@ -0,0 +1,173 @@
+#  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.
+"""Contains classes for MPNetForTokenClassification."""
+
+from sparknlp.common import *
+
+
+class MPNetForTokenClassification(AnnotatorModel,
+                                       HasCaseSensitiveProperties,
+                                       HasBatchedAnnotate,
+                                       HasEngine,
+                                       HasMaxSentenceLengthLimit):
+    """MPNetForTokenClassification can load XLM-RoBERTa Models with a token
+    classification head on top (a linear layer on top of the hidden-states
+    output) e.g. for Named-Entity-Recognition (NER) tasks.
+
+    Pretrained models can be loaded with :meth:`.pretrained` of the companion
+    object:
+
+    >>> token_classifier = MPNetForTokenClassification.pretrained() \\
+    ...     .setInputCols(["token", "document"]) \\
+    ...     .setOutputCol("label")
+    The default model is ``"mpnet_base_token_classifier"``, if no
+    name is provided.
+
+    For available pretrained models please see the `Models Hub
+    <https://sparknlp.org/models?task=Named+Entity+Recognition>`__.
+    To see which models are compatible and how to import them see
+    `Import Transformers into Spark NLP 🚀
+    <https://github.com/JohnSnowLabs/spark-nlp/discussions/5669>`_.
+
+    ====================== ======================
+    Input Annotation types Output Annotation type
+    ====================== ======================
+    ``DOCUMENT, TOKEN``    ``NAMED_ENTITY``
+    ====================== ======================
+
+    Parameters
+    ----------
+    batchSize
+        Batch size. Large values allows faster processing but requires more
+        memory, by default 8
+    caseSensitive
+        Whether to ignore case in tokens for embeddings matching, by default
+        True
+    configProtoBytes
+        ConfigProto from tensorflow, serialized into byte array.
+    maxSentenceLength
+        Max sentence length to process, by default 128
+
+    Examples
+    --------
+    >>> import sparknlp
+    >>> from sparknlp.base import *
+    >>> from sparknlp.annotator import *
+    >>> from pyspark.ml import Pipeline
+    >>> documentAssembler = DocumentAssembler() \\
+    ...     .setInputCol("text") \\
+    ...     .setOutputCol("document")
+    >>> tokenizer = Tokenizer() \\
+    ...     .setInputCols(["document"]) \\
+    ...     .setOutputCol("token")
+    >>> tokenClassifier = MPNetForTokenClassification.pretrained() \\
+    ...     .setInputCols(["token", "document"]) \\
+    ...     .setOutputCol("label") \\
+    ...     .setCaseSensitive(True)
+    >>> pipeline = Pipeline().setStages([
+    ...     documentAssembler,
+    ...     tokenizer,
+    ...     tokenClassifier
+    ... ])
+    >>> data = spark.createDataFrame([["John Lenon was born in London and lived in Paris. My name is Sarah and I live in London"]]).toDF("text")
+    >>> result = pipeline.fit(data).transform(data)
+    >>> result.select("label.result").show(truncate=False)
+    +------------------------------------------------------------------------------------+
+    |result                                                                              |
+    +------------------------------------------------------------------------------------+
+    |[B-PER, I-PER, O, O, O, B-LOC, O, O, O, B-LOC, O, O, O, O, B-PER, O, O, O, O, B-LOC]|
+    +------------------------------------------------------------------------------------+
+    """
+    name = "MPNetForTokenClassification"
+
+    inputAnnotatorTypes = [AnnotatorType.DOCUMENT, AnnotatorType.TOKEN]
+
+    outputAnnotatorType = AnnotatorType.NAMED_ENTITY
+
+    configProtoBytes = Param(Params._dummy(),
+                             "configProtoBytes",
+                             "ConfigProto from tensorflow, serialized into byte array. Get with config_proto.SerializeToString()",
+                             TypeConverters.toListInt)
+
+    def getClasses(self):
+        """
+        Returns labels used to train this model
+        """
+        return self._call_java("getClasses")
+
+    def setConfigProtoBytes(self, b):
+        """Sets configProto from tensorflow, serialized into byte array.
+
+        Parameters
+        ----------
+        b : List[int]
+            ConfigProto from tensorflow, serialized into byte array
+        """
+        return self._set(configProtoBytes=b)
+
+    @keyword_only
+    def __init__(self, classname="com.johnsnowlabs.nlp.annotators.classifier.dl.MPNetForTokenClassification",
+                 java_model=None):
+        super(MPNetForTokenClassification, self).__init__(
+            classname=classname,
+            java_model=java_model
+        )
+        self._setDefault(
+            batchSize=8,
+            maxSentenceLength=128,
+            caseSensitive=True
+        )
+
+    @staticmethod
+    def loadSavedModel(folder, spark_session):
+        """Loads a locally saved model.
+
+        Parameters
+        ----------
+        folder : str
+            Folder of the saved model
+        spark_session : pyspark.sql.SparkSession
+            The current SparkSession
+
+        Returns
+        -------
+        XlmRoBertaForTokenClassification
+            The restored model
+        """
+        from sparknlp.internal import _MPNetForTokenClassifierLoader
+        jModel = _MPNetForTokenClassifierLoader(folder, spark_session._jsparkSession)._java_obj
+        return MPNetForTokenClassification(java_model=jModel)
+
+    @staticmethod
+    def pretrained(name="mpnet_base_token_classifier", lang="en", remote_loc=None):
+        """Downloads and loads a pretrained model.
+
+        Parameters
+        ----------
+        name : str, optional
+            Name of the pretrained model, by default
+            "mpnet_base_token_classifier"
+        lang : str, optional
+            Language of the pretrained model, by default "en"
+        remote_loc : str, optional
+            Optional remote address of the resource, by default None. Will use
+            Spark NLPs repositories otherwise.
+
+        Returns
+        -------
+        XlmRoBertaForTokenClassification
+            The restored model
+        """
+        from sparknlp.pretrained import ResourceDownloader
+        return ResourceDownloader.downloadModel(MPNetForTokenClassification, name, lang, remote_loc)

sparknlp/annotator/classifier_dl/xlm_roberta_for_token_classification.py

@@ -31,7 +31,7 @@ class XlmRoBertaForTokenClassification(AnnotatorModel,
     >>> token_classifier = XlmRoBertaForTokenClassification.pretrained() \\
     ...     .setInputCols(["token", "document"]) \\
     ...     .setOutputCol("label")
-    The default model is ``"xlm_roberta_base_token_classifier_conll03"``, if no
+    The default model is ``"mpnet_base_token_classifier"``, if no
     name is provided.
 
     For available pretrained models please see the `Models Hub
@@ -150,14 +150,14 @@ class XlmRoBertaForTokenClassification(AnnotatorModel,
         return XlmRoBertaForTokenClassification(java_model=jModel)
 
     @staticmethod
-    def pretrained(name="xlm_roberta_base_token_classifier_conll03", lang="en", remote_loc=None):
+    def pretrained(name="mpnet_base_token_classifier", lang="en", remote_loc=None):
         """Downloads and loads a pretrained model.
 
         Parameters
         ----------
         name : str, optional
             Name of the pretrained model, by default
-            "xlm_roberta_base_token_classifier_conll03"
+            "mpnet_base_token_classifier"
         lang : str, optional
             Language of the pretrained model, by default "en"
         remote_loc : str, optional

sparknlp/annotator/embeddings/__init__.py

@@ -36,3 +36,4 @@ from sparknlp.annotator.embeddings.xlm_roberta_embeddings import *
 from sparknlp.annotator.embeddings.xlm_roberta_sentence_embeddings import *
 from sparknlp.annotator.embeddings.xlnet_embeddings import *
 from sparknlp.annotator.embeddings.bge_embeddings import *
+from sparknlp.annotator.embeddings.uae_embeddings import *

sparknlp/annotator/embeddings/bert_embeddings.py

@@ -164,7 +164,7 @@ class BertEmbeddings(AnnotatorModel,
         )
 
     @staticmethod
-    def loadSavedModel(folder, spark_session):
+    def loadSavedModel(folder, spark_session, use_openvino=False):
         """Loads a locally saved model.
 
         Parameters
@@ -173,6 +173,8 @@ class BertEmbeddings(AnnotatorModel,
             Folder of the saved model
         spark_session : pyspark.sql.SparkSession
             The current SparkSession
+        use_openvino: bool
+            Use OpenVINO backend
 
         Returns
         -------
@@ -180,7 +182,7 @@ class BertEmbeddings(AnnotatorModel,
             The restored model
         """
         from sparknlp.internal import _BertLoader
-        jModel = _BertLoader(folder, spark_session._jsparkSession)._java_obj
+        jModel = _BertLoader(folder, spark_session._jsparkSession, use_openvino)._java_obj
         return BertEmbeddings(java_model=jModel)
 
     @staticmethod

sparknlp/annotator/embeddings/bert_sentence_embeddings.py

@@ -180,7 +180,7 @@ class BertSentenceEmbeddings(AnnotatorModel,
         )
 
     @staticmethod
-    def loadSavedModel(folder, spark_session):
+    def loadSavedModel(folder, spark_session, use_openvino=False):
         """Loads a locally saved model.
 
         Parameters
@@ -189,6 +189,8 @@ class BertSentenceEmbeddings(AnnotatorModel,
             Folder of the saved model
         spark_session : pyspark.sql.SparkSession
             The current SparkSession
+        use_openvino: bool
+            Use OpenVINO backend
 
         Returns
         -------
@@ -196,7 +198,7 @@ class BertSentenceEmbeddings(AnnotatorModel,
             The restored model
         """
         from sparknlp.internal import _BertSentenceLoader
-        jModel = _BertSentenceLoader(folder, spark_session._jsparkSession)._java_obj
+        jModel = _BertSentenceLoader(folder, spark_session._jsparkSession, use_openvino)._java_obj
         return BertSentenceEmbeddings(java_model=jModel)
 
     @staticmethod

sparknlp/annotator/embeddings/bge_embeddings.py

@@ -26,6 +26,8 @@ class BGEEmbeddings(AnnotatorModel,
 
    BGE, or BAAI General Embeddings, a model that can map any text to a low-dimensional dense 
   vector which can be used for tasks like retrieval, classification, clustering, or semantic search.
+
+  Note that this annotator is only supported for Spark Versions 3.4 and up.
   
   Pretrained models can be loaded with `pretrained` of the companion object:
 

sparknlp/annotator/embeddings/e5_embeddings.py

@@ -25,6 +25,8 @@ class E5Embeddings(AnnotatorModel,
     """Sentence embeddings using E5.
 
     E5, a weakly supervised text embedding model that can generate text embeddings tailored to any task (e.g., classification, retrieval, clustering, text evaluation, etc.)
+    Note that this annotator is only supported for Spark Versions 3.4 and up.
+
     Pretrained models can be loaded with :meth:`.pretrained` of the companion
     object:
 
@@ -149,7 +151,7 @@ class E5Embeddings(AnnotatorModel,
         )
 
     @staticmethod
-    def loadSavedModel(folder, spark_session):
+    def loadSavedModel(folder, spark_session, use_openvino=False):
         """Loads a locally saved model.
 
         Parameters
@@ -158,6 +160,8 @@ class E5Embeddings(AnnotatorModel,
             Folder of the saved model
         spark_session : pyspark.sql.SparkSession
             The current SparkSession
+        use_openvino : bool
+            Use OpenVINO backend
 
         Returns
         -------
@@ -165,7 +169,7 @@ class E5Embeddings(AnnotatorModel,
             The restored model
         """
         from sparknlp.internal import _E5Loader
-        jModel = _E5Loader(folder, spark_session._jsparkSession)._java_obj
+        jModel = _E5Loader(folder, spark_session._jsparkSession, use_openvino)._java_obj
         return E5Embeddings(java_model=jModel)
 
     @staticmethod

sparknlp/annotator/embeddings/mpnet_embeddings.py

@@ -28,6 +28,8 @@ class MPNetEmbeddings(AnnotatorModel,
     to inherit the advantages of masked language modeling and permuted language modeling for
     natural language understanding.
 
+    Note that this annotator is only supported for Spark Versions 3.4 and up.
+
     Pretrained models can be loaded with :meth:`.pretrained` of the companion
     object:
 

sparknlp/annotator/embeddings/roberta_embeddings.py

@@ -181,7 +181,7 @@ class RoBertaEmbeddings(AnnotatorModel,
         )
 
     @staticmethod
-    def loadSavedModel(folder, spark_session):
+    def loadSavedModel(folder, spark_session, use_openvino=False):
         """Loads a locally saved model.
 
         Parameters
@@ -190,6 +190,8 @@ class RoBertaEmbeddings(AnnotatorModel,
             Folder of the saved model
         spark_session : pyspark.sql.SparkSession
             The current SparkSession
+        use_openvino: bool
+            Use OpenVINO backend
 
         Returns
         -------
@@ -197,7 +199,7 @@ class RoBertaEmbeddings(AnnotatorModel,
             The restored model
         """
         from sparknlp.internal import _RoBertaLoader
-        jModel = _RoBertaLoader(folder, spark_session._jsparkSession)._java_obj
+        jModel = _RoBertaLoader(folder, spark_session._jsparkSession, use_openvino)._java_obj
         return RoBertaEmbeddings(java_model=jModel)
 
     @staticmethod

sparknlp/annotator/embeddings/uae_embeddings.py

@@ -0,0 +1,211 @@
+#  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.
+"""Contains classes for UAEEmbeddings."""
+
+from sparknlp.common import *
+
+
+class UAEEmbeddings(AnnotatorModel,
+                    HasEmbeddingsProperties,
+                    HasCaseSensitiveProperties,
+                    HasStorageRef,
+                    HasBatchedAnnotate,
+                    HasMaxSentenceLengthLimit):
+    """Sentence embeddings using Universal AnglE Embedding (UAE).
+
+    UAE is a novel angle-optimized text embedding model, designed to improve semantic textual
+    similarity tasks, which are crucial for Large Language Model (LLM) applications. By
+    introducing angle optimization in a complex space, AnglE effectively mitigates saturation of
+    the cosine similarity function.
+
+    Pretrained models can be loaded with :meth:`.pretrained` of the companion
+    object:
+
+    >>> embeddings = UAEEmbeddings.pretrained() \\
+    ...     .setInputCols(["document"]) \\
+    ...     .setOutputCol("UAE_embeddings")
+
+
+    The default model is ``"uae_large_v1"``, if no name is provided.
+
+    For available pretrained models please see the
+    `Models Hub <https://sparknlp.org/models?q=UAE>`__.
+
+
+    ====================== ======================
+    Input Annotation types Output Annotation type
+    ====================== ======================
+    ``DOCUMENT``            ``SENTENCE_EMBEDDINGS``
+    ====================== ======================
+
+    Parameters
+    ----------
+    batchSize
+        Size of every batch , by default 8
+    dimension
+        Number of embedding dimensions, by default 768
+    caseSensitive
+        Whether to ignore case in tokens for embeddings matching, by default False
+    maxSentenceLength
+        Max sentence length to process, by default 512
+    configProtoBytes
+        ConfigProto from tensorflow, serialized into byte array.
+
+    References
+    ----------
+
+    `AnglE-optimized Text Embeddings <https://arxiv.org/abs/2309.12871>`__
+    `UAE Github Repository <https://github.com/baochi0212/uae-embedding>`__
+
+    **Paper abstract**
+
+    *High-quality text embedding is pivotal in improving semantic textual similarity (STS) tasks,
+    which are crucial components in Large Language Model (LLM) applications. However, a common
+    challenge existing text embedding models face is the problem of vanishing gradients, primarily
+    due to their reliance on the cosine function in the optimization objective, which has
+    saturation zones. To address this issue, this paper proposes a novel angle-optimized text
+    embedding model called AnglE. The core idea of AnglE is to introduce angle optimization in a
+    complex space. This novel approach effectively mitigates the adverse effects of the saturation
+    zone in the cosine function, which can impede gradient and hinder optimization processes. To
+    set up a comprehensive STS evaluation, we experimented on existing short-text STS datasets and
+    a newly collected long-text STS dataset from GitHub Issues. Furthermore, we examine
+    domain-specific STS scenarios with limited labeled data and explore how AnglE works with
+    LLM-annotated data. Extensive experiments were conducted on various tasks including short-text
+    STS, long-text STS, and domain-specific STS tasks. The results show that AnglE outperforms the
+    state-of-the-art (SOTA) STS models that ignore the cosine saturation zone. These findings
+    demonstrate the ability of AnglE to generate high-quality text embeddings and the usefulness
+    of angle optimization in STS.*
+
+    Examples
+    --------
+    >>> import sparknlp
+    >>> from sparknlp.base import *
+    >>> from sparknlp.annotator import *
+    >>> from pyspark.ml import Pipeline
+    >>> documentAssembler = DocumentAssembler() \\
+    ...     .setInputCol("text") \\
+    ...     .setOutputCol("document")
+    >>> embeddings = UAEEmbeddings.pretrained() \\
+    ...     .setInputCols(["document"]) \\
+    ...     .setOutputCol("embeddings")
+    >>> embeddingsFinisher = EmbeddingsFinisher() \\
+    ...     .setInputCols("embeddings") \\
+    ...     .setOutputCols("finished_embeddings") \\
+    ...     .setOutputAsVector(True)
+    >>> pipeline = Pipeline().setStages([
+    ...     documentAssembler,
+    ...     embeddings,
+    ...     embeddingsFinisher
+    ... ])
+    >>> data = spark.createDataFrame([["hello world", "hello moon"]]).toDF("text")
+    >>> result = pipeline.fit(data).transform(data)
+    >>> result.selectExpr("explode(finished_embeddings) as result").show(5, 80)
+    +--------------------------------------------------------------------------------+
+    |                                                                          result|
+    +--------------------------------------------------------------------------------+
+    |[0.50387806, 0.5861606, 0.35129607, -0.76046336, -0.32446072, -0.117674336, 0...|
+    |[0.6660665, 0.961762, 0.24854276, -0.1018044, -0.6569202, 0.027635604, 0.1915...|
+    +--------------------------------------------------------------------------------+
+    """
+
+    name = "UAEEmbeddings"
+
+    inputAnnotatorTypes = [AnnotatorType.DOCUMENT]
+
+    outputAnnotatorType = AnnotatorType.SENTENCE_EMBEDDINGS
+    poolingStrategy = Param(Params._dummy(),
+                            "poolingStrategy",
+                            "Pooling strategy to use for sentence embeddings",
+                            TypeConverters.toString)
+
+    def setPoolingStrategy(self, value):
+        """Pooling strategy to use for sentence embeddings.
+
+        Available pooling strategies for sentence embeddings are:
+          - `"cls"`: leading `[CLS]` token
+          - `"cls_avg"`: leading `[CLS]` token + mean of all other tokens
+          - `"last"`: embeddings of the last token in the sequence
+          - `"avg"`: mean of all tokens
+          - `"max"`: max of all embedding features of the entire token sequence
+          - `"int"`: An integer number, which represents the index of the token to use as the
+            embedding
+
+        Parameters
+        ----------
+        value : str
+            Pooling strategy to use for sentence embeddings
+        """
+
+        valid_strategies = {"cls", "cls_avg", "last", "avg", "max"}
+        if value in valid_strategies or value.isdigit():
+            return self._set(poolingStrategy=value)
+        else:
+            raise ValueError(f"Invalid pooling strategy: {value}. "
+                             f"Valid strategies are: {', '.join(self.valid_strategies)} or an integer.")
+
+    @keyword_only
+    def __init__(self, classname="com.johnsnowlabs.nlp.embeddings.UAEEmbeddings", java_model=None):
+        super(UAEEmbeddings, self).__init__(
+            classname=classname,
+            java_model=java_model
+        )
+        self._setDefault(
+            dimension=1024,
+            batchSize=8,
+            maxSentenceLength=512,
+            caseSensitive=False,
+            poolingStrategy="cls"
+        )
+
+    @staticmethod
+    def loadSavedModel(folder, spark_session):
+        """Loads a locally saved model.
+
+        Parameters
+        ----------
+        folder : str
+            Folder of the saved model
+        spark_session : pyspark.sql.SparkSession
+            The current SparkSession
+
+        Returns
+        -------
+        UAEEmbeddings
+            The restored model
+        """
+        from sparknlp.internal import _UAEEmbeddingsLoader
+        jModel = _UAEEmbeddingsLoader(folder, spark_session._jsparkSession)._java_obj
+        return UAEEmbeddings(java_model=jModel)
+
+    @staticmethod
+    def pretrained(name="uae_large_v1", lang="en", remote_loc=None):
+        """Downloads and loads a pretrained model.
+
+        Parameters
+        ----------
+        name : str, optional
+            Name of the pretrained model, by default "UAE_small"
+        lang : str, optional
+            Language of the pretrained model, by default "en"
+        remote_loc : str, optional
+            Optional remote address of the resource, by default None. Will use
+            Spark NLPs repositories otherwise.
+
+        Returns
+        -------
+        UAEEmbeddings
+            The restored model
+        """
+        from sparknlp.pretrained import ResourceDownloader
+        return ResourceDownloader.downloadModel(UAEEmbeddings, name, lang, remote_loc)

sparknlp/annotator/embeddings/xlm_roberta_embeddings.py

@@ -181,7 +181,7 @@ class XlmRoBertaEmbeddings(AnnotatorModel,
         )
 
     @staticmethod
-    def loadSavedModel(folder, spark_session):
+    def loadSavedModel(folder, spark_session, use_openvino=False):
         """Loads a locally saved model.
 
         Parameters
@@ -190,6 +190,8 @@ class XlmRoBertaEmbeddings(AnnotatorModel,
             Folder of the saved model
         spark_session : pyspark.sql.SparkSession
             The current SparkSession
+        use_openvino: bool
+            Use OpenVINO backend
 
         Returns
         -------
@@ -197,7 +199,7 @@ class XlmRoBertaEmbeddings(AnnotatorModel,
             The restored model
         """
         from sparknlp.internal import _XlmRoBertaLoader
-        jModel = _XlmRoBertaLoader(folder, spark_session._jsparkSession)._java_obj
+        jModel = _XlmRoBertaLoader(folder, spark_session._jsparkSession, use_openvino)._java_obj
         return XlmRoBertaEmbeddings(java_model=jModel)
 
     @staticmethod