diff --git a/dialogy/plugins/text/duckling_plugin/__init__.py b/dialogy/plugins/text/duckling_plugin/__init__.py index 5b09064c..b1a6c5ba 100644 --- a/dialogy/plugins/text/duckling_plugin/__init__.py +++ b/dialogy/plugins/text/duckling_plugin/__init__.py @@ -597,11 +597,12 @@ def __create_req_body( :rtype: Dict[str, Any] """ dimensions = self.dimensions - self.activate_latent_entities = use_latent or self.activate_latent_entities + activate_latent_entities = use_latent + activate_latent_entities = ( - self.activate_latent_entities - if isinstance(self.activate_latent_entities, bool) - else self.activate_latent_entities() + activate_latent_entities + if isinstance(activate_latent_entities, bool) + else activate_latent_entities() ) payload = { diff --git a/docs/.buildinfo b/docs/.buildinfo index 63da9ce5..52639e68 100644 --- a/docs/.buildinfo +++ b/docs/.buildinfo @@ -1,4 +1,4 @@ # Sphinx build info version 1 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done. -config: c49aa7e37a5eef9d9b5d7193dbe5cf54 +config: 7d590465bce22dcff23310e826e335c2 tags: 645f666f9bcd5a90fca523b33c5a78b7 diff --git a/docs/_modules/dialogy/base/entity_extractor.html b/docs/_modules/dialogy/base/entity_extractor.html index 7fed3972..f1803cfc 100644 --- a/docs/_modules/dialogy/base/entity_extractor.html +++ b/docs/_modules/dialogy/base/entity_extractor.html @@ -5,7 +5,7 @@ - dialogy.base.entity_extractor - dialogy 0.9.1 documentation + dialogy.base.entity_extractor - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
-
dialogy 0.9.1 documentation
 +
dialogy 0.9.4 documentation
@@ -147,7 +147,7 @@
- dialogy 0.9.1 documentation + dialogy 0.9.4 documentation
diff --git a/docs/_modules/dialogy/cli.html b/docs/_modules/dialogy/cli.html index 5fab0739..43680627 100644 --- a/docs/_modules/dialogy/cli.html +++ b/docs/_modules/dialogy/cli.html @@ -5,7 +5,7 @@ - dialogy.cli - dialogy 0.9.1 documentation + dialogy.cli - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
-
dialogy 0.9.1 documentation
 +
dialogy 0.9.4 documentation
@@ -147,7 +147,7 @@
- dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
  • Input
  • Output
  • Workflow
    • -
  • Base Plugin
    • +
  • Plugin
    • +
  • Intent
    • +
  • Entity
    • +
  • Slots

    • Types

    Plugins

    Misc

    Types

    Plugins

    Misc

    Types

    Plugins

    Misc

      @@ -221,9 +223,9 @@

      Source code for dialogy.cli.workflow

       from typing import Any
 
 import pandas as pd
-from loguru import logger
 
 import dialogy.constants as const
+from dialogy.utils import logger
 from dialogy.workflow import Workflow
 
 
@@ -258,6 +260,7 @@ 
      Source code for dialogy.cli.workflow
           CLI entry point for workflows.
 
     args contain the following attributes:
+    ----
 
     module      - We expect the Workflow or its subclass to be available at this path.
     fn          - A function that provides a Workflow or its subclass' instance.
@@ -310,5 +313,7 @@ 
      Source code for dialogy.cli.workflow
           
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/constants.html b/docs/_modules/dialogy/constants.html
index 3c134a65..7e51be49 100644
--- a/docs/_modules/dialogy/constants.html
+++ b/docs/_modules/dialogy/constants.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.constants - dialogy 0.9.3 documentation
+        dialogy.constants - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.3 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -161,10 +161,12 @@
 
    • Output
      • 
 
    • Workflow
      • 
 
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

      @@ -219,13 +220,14 @@

      Source code for dialogy.constants

       """Module provides Constants to be used throughtout the project."""
 
-from locale import DAY_1
-
 DEFAULT_NAMESPACE = "vernacular-ai"
 DEFAULT_PROJECT_TEMPLATE = "dialogy-template-simple-transformers"
 
 PLUGINS = "plugins"
+PREPROCESSORS = "preprocessors"
+POSTPROCESSORS = "postprocessors"
 DEBUG = "debug"
+WORKFLOW_PUBLIC_PROPERTIES = [PREPROCESSORS, POSTPROCESSORS]
 S_INTENT_OOS = "_oos_"
 TRANSCRIPT = "transcript"
 AM_SCORE = "am_score"
@@ -234,68 +236,38 @@ 
      Source code for dialogy.constants
       TO = "to"
 FROM = "from"
 LOCK = "lock"
-NAME = "name"
-SLOTS = "slots"
 
-DUCKLING_ENTITY_KEYS = ("body", "dim", "end", "latent", "start", "value")
-DUCKLING_TIME_VALUES_ENTITY_KEYS = ("grain", "type", "value", "values")
-DUCKLING_TIME_INTERVAL_ENTITY_KEYS = (
-    ("from", "type", "values"),
-    ("to", "type", "values"),
-    ("from", "to", "type", "values"),
-)
 
-# Time Entity types
-DATE = "date"
-TIME = "time"
-DATETIME = "datetime"
-HOUR = "hour"
-DAY = "day"
-MONTH = "month"
-
-PEOPLE = "people"
-NUMBER = "number"
-AMOUNT_OF_MONEY = "amount-of-money"
-DURATION = "duration"
-CREDIT_CARD_NUMBER = "credit-card-number"
+
      [docs]class EntityKeys:
+    BODY = "body"
+    DIM = "dim"
+    END = "end"
+    FROM = "from"
+    GRAIN = "grain"
+    INTERVAL = "interval"
+    DURATION = "duration"
+    LATENT = "latent"
+    RANGE = "range"
+    START = "start"
+    TO = "to"
+    TYPE = "type"
+    UNIT = "unit"
+    VALUE = "value"
+    VALUES = "values"
+    ORIGIN = "origin"
+    NORMALIZED = "normalized"
+    SLOT_NAMES = "slot_names"
+    ALTERNATIVE_INDEX = "alternative_index"
+    ALTERNATIVE_INDICES = "alternative_indices"
      
 
-FUTURE = "future"
-PAST = "past"
-ANY = "__any__"
-
-DUCKLING_DIMS = [PEOPLE, NUMBER, AMOUNT_OF_MONEY, DURATION, CREDIT_CARD_NUMBER, TIME]
-
-BODY = "body"
-DIM = "dim"
-END = "end"
-FROM = "from"
-GRAIN = "grain"
-INTERVAL = "interval"
-DURATION = "duration"
-LATENT = "latent"
-RANGE = "range"
-START = "start"
-TO = "to"
-TYPE = "type"
-ENTITY_TYPE = "entity_type"
-UNIT = "unit"
-VALUE = "value"
-VALUES = "values"
-ORIGIN = "origin"
-ISSUER = "issuer"
-NORMALIZED = "normalized"
-ALTERNATIVE_INDEX = "alternative_index"
-ALTERNATIVE_INDICES = "alternative_indices"
-CREDIT_CARD_NUMBER = "credit-card-number"
 
-SLOT_NAMES = "slot_names"
 SKIP_ENTITY_ATTRS = [
-    DIM,
-    VALUES,
-    LATENT,
-    ORIGIN,
-    SLOT_NAMES,
-    ALTERNATIVE_INDICES,
+    EntityKeys.DIM,
+    EntityKeys.VALUES,
+    EntityKeys.LATENT,
+    EntityKeys.ORIGIN,
+    EntityKeys.SLOT_NAMES,
+    EntityKeys.ALTERNATIVE_INDICES,
 ]
 
 # This section is needed for dialogy.types.entities.*
@@ -307,32 +279,27 @@ 
      Source code for dialogy.constants
       # instantiate the Entity. We use this mapping to ensure that the
 # values belong to expected types.
 BASE_ENTITY_PROPS = [
-    ([RANGE], dict),
-    ([RANGE, START], int),
-    ([RANGE, END], int),
-    ([BODY], str),
-    ([VALUES], list),
-    ([DIM], str),
-    ([LATENT], bool),
-]
-
-CREDIT_CARD_NUMBER_PROPS = [
-    ([RANGE], dict),
-    ([RANGE, START], int),
-    ([RANGE, END], int),
-    ([BODY], str),
-    ([VALUE], dict),
-    ([DIM], str),
-    ([LATENT], bool),
+    ([EntityKeys.RANGE], dict),
+    ([EntityKeys.RANGE, EntityKeys.START], int),
+    ([EntityKeys.RANGE, EntityKeys.END], int),
+    ([EntityKeys.BODY], str),
+    ([EntityKeys.VALUES], list),
+    ([EntityKeys.DIM], str),
+    ([EntityKeys.LATENT], bool),
 ]
 
-TIME_ENTITY_PROPS = BASE_ENTITY_PROPS + [([GRAIN], str)]
-PEOPLE_ENTITY_PROPS = BASE_ENTITY_PROPS + [([UNIT], str)]
+TIME_ENTITY_PROPS = BASE_ENTITY_PROPS + [([EntityKeys.GRAIN], str)]
+PEOPLE_ENTITY_PROPS = BASE_ENTITY_PROPS + [([EntityKeys.UNIT], str)]
 
 # Time related constants
 DATE_UNITS = ["day", "week", "month", "quarter", "year"]
 TIME_UNITS = ["hour", "minute", "second"]
 
+# Time Entity types
+DATE = "date"
+TIME = "time"
+DATETIME = "datetime"
+
 
 # This section covers signals tuple indices
 
      [docs]class SIGNAL:
@@ -353,14 +320,12 @@ 
      Source code for dialogy.constants
       
 # XLMR
 ID = "id"
-IDX = "idx"
 DATA = "data"
-STATE = "state"
 CLASSIFICATION_INPUT = "classification_input"
 XLMR_MODULE = "simpletransformers.classification"
 XLMR_MULTI_CLASS_MODEL = "ClassificationModel"
-XLMR_MODEL = "xlmroberta"
-XLMR_MODEL_TIER = "xlm-roberta-base"
+XLMR_MODEL = "roberta"
+XLMR_MODEL_TIER = "roberta-base"
 INTENT = "intent"
 INTENTS = f"{INTENT}s"
 LABELS = "labels"
@@ -369,27 +334,6 @@ 
      Source code for dialogy.constants
       LABELENCODER_FILE = "labelencoder.pkl"
 INVALID_TOKENS = ["<UNK>", "<PAD>", "<SOS>", "<EOS>"]
 
-# MLP
-MLPMODEL_FILE = "mlpmodelpipeline.joblib"
-NUM_TRAIN_EPOCHS = "num_train_epochs"
-MLP_DEFAULT_TRAIN_EPOCHS = 25
-DEFAULT_NGRAM = (1, 1)
-TFIDF_LOWERCASE = False
-STOPWORDS = None
-MLP_RANDOMSTATE = 1
-
-TFIDF = "tfidf"
-MLP = "mlp"
-
-# Hyperparams For MLP
-USE_GRIDSEARCH = "gridsearch_hyperparams"
-PARAMS = "params"
-GRID_SCORETYPE = "weighted"
-CV = "cv"
-VERBOSE_LEVEL = "verbose_level"
-NGRAM_RANGE = "ngram_range"
-GRIDSEARCH_WORKERS = -1  # -1 means use all available cores
-
 # CLI Commands
 TRAIN = "train"
 TEST = "test"
@@ -401,8 +345,6 @@ 
      Source code for dialogy.constants
       CANONICALIZED_TRANSCRIPTS = "canonicalized_transcripts"
 REFERENCE_TIME = "reftime"
 ENTITY_COLUMN = "entities"
-
-FUZZY_DP = "fuzzy_dp"
 
      
         
       
      
diff --git a/docs/_modules/dialogy/plugins/text/calibration/xgb.html b/docs/_modules/dialogy/plugins/text/calibration/xgb.html
index d7d5e578..a00e87db 100644
--- a/docs/_modules/dialogy/plugins/text/calibration/xgb.html
+++ b/docs/_modules/dialogy/plugins/text/calibration/xgb.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.plugins.text.calibration.xgb - dialogy 0.9.1 documentation
+        dialogy.plugins.text.calibration.xgb - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.1 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.1 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -160,11 +160,13 @@
 
    • Input
      • 
 
    • Output
      • 
 
    • Workflow
      • 
-
    • Base Plugin
      • 
+
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

    Misc

      @@ -231,16 +233,15 @@

      Source code for dialogy.plugins.text.calibration.xgb

      import jiwer import numpy as np import pandas as pd -from loguru import logger from sklearn.base import BaseEstimator, TransformerMixin from sklearn.feature_extraction.text import TfidfVectorizer from tqdm import tqdm from xgboost import XGBRegressor from dialogy import constants as const -from dialogy.base import Guard, Input, Output, Plugin -from dialogy.types import Transcript, Utterance, utterances -from dialogy.utils import normalize +from dialogy.base.plugin import Plugin, PluginFn +from dialogy.types import Transcript, Utterance +from dialogy.utils import logger, normalize
      [docs]class FeatureExtractor(BaseEstimator, TransformerMixin): @@ -302,8 +303,6 @@

      Source code for dialogy.plugins.text.calibration.xgb

      [docs]class CalibrationModel(Plugin): """ - .. _calibration_plugin: - This plugin provides a calibration model that sits between ASR and SLU. It trains a model that learn to classify alternatives from the text and AM, LM score. Bad alternatives are removed before training SLU and during @@ -312,9 +311,9 @@

      Source code for dialogy.plugins.text.calibration.xgb

      def __init__( self, + access: Optional[PluginFn], + mutate: Optional[PluginFn], threshold: float, - dest: Optional[str] = None, - guards: Optional[List[Guard]] = None, debug: bool = False, input_column: str = const.ALTERNATIVES, output_column: Optional[str] = const.ALTERNATIVES, @@ -322,8 +321,8 @@

      Source code for dialogy.plugins.text.calibration.xgb

      model_name: str = "calibration.pkl", ) -> None: super().__init__( - dest=dest, - guards=guards, + access, + mutate, debug=debug, input_column=input_column, output_column=output_column, @@ -400,9 +399,8 @@

      Source code for dialogy.plugins.text.calibration.xgb

      ) return training_data_
      -
      [docs] def inference( - self, transcripts: List[str], utterances: List[Utterance] - ) -> List[str]: +
      [docs] def inference(self, utterances: List[Utterance]) -> List[str]: + transcripts: List[Transcript] = normalize(utterances) transcript_lengths: List[int] = [ len(transcript.split()) for transcript in transcripts ] @@ -415,17 +413,15 @@

      Source code for dialogy.plugins.text.calibration.xgb

      # a classifier's prediction to a fallback label. # If the transcripts have less than WORD_THRESHOLD words, we will always predict the fallback label. if average_word_count <= const.WORD_THRESHOLD: - return transcripts + return normalize(utterances) return normalize(self.filter_asr_output(utterances))
      [docs] def save(self, fname: str) -> None: pickle.dump(self, open(fname, "wb"))
      -
      [docs] def utility(self, input: Input, _: Output) -> Any: - return self.inference( - input.transcripts, input.utterances - ) # pylint: disable=no-value-for-parameter
      +
      [docs] def utility(self, *args: Any) -> Any: + return self.inference(*args) # pylint: disable=no-value-for-parameter
      [docs] def validate(self, df: pd.DataFrame) -> bool: """ @@ -480,5 +476,7 @@

      Source code for dialogy.plugins.text.calibration.xgb

      + + \ No newline at end of file diff --git a/docs/_modules/dialogy/plugins/text/canonicalization.html b/docs/_modules/dialogy/plugins/text/canonicalization.html index 78e620ee..bc4011a2 100644 --- a/docs/_modules/dialogy/plugins/text/canonicalization.html +++ b/docs/_modules/dialogy/plugins/text/canonicalization.html @@ -5,7 +5,7 @@ - dialogy.plugins.text.canonicalization - dialogy 0.9.1 documentation + dialogy.plugins.text.canonicalization - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.1 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
    • Input
    • Output
    • Workflow
      • -
    • Base Plugin
      • +
    • Plugin
      • +
    • Intent
      • +
    • Entity
      • +
    • Slots

    Types

    Plugins

    Misc

      @@ -223,17 +225,16 @@

      Source code for dialogy.plugins.text.canonicalization

      from typing import Any, Callable, List, Optional import pandas as pd -from loguru import logger from tqdm import tqdm import dialogy.constants as const -from dialogy.base import Guard, Input, Output, Plugin -from dialogy.types import BaseEntity -from dialogy.utils import normalize +from dialogy.base.plugin import Plugin +from dialogy.types import BaseEntity, plugin +from dialogy.utils import logger, normalize
      [docs]def get_entity_type(entity: BaseEntity) -> str: - return f"<{entity.entity_type}>"
      + return f"<{entity.type}>"
      [docs]class CanonicalizationPlugin(Plugin): @@ -246,8 +247,8 @@

      Source code for dialogy.plugins.text.canonicalization

      serializer: Callable[[BaseEntity], str] = get_entity_type, mask: str = "MASK", mask_tokens: Optional[List[str]] = None, - dest: Optional[str] = None, - guards: Optional[List[Guard]] = None, + access: Optional[plugin.PluginFn] = None, + mutate: Optional[plugin.PluginFn] = None, entity_column: str = const.ENTITY_COLUMN, input_column: str = const.ALTERNATIVES, output_column: Optional[str] = None, @@ -256,8 +257,8 @@

      Source code for dialogy.plugins.text.canonicalization

      debug: bool = False, ) -> None: super().__init__( - dest=dest, - guards=guards, + access, + mutate, debug=debug, use_transform=use_transform, input_column=input_column, @@ -282,8 +283,8 @@

      Source code for dialogy.plugins.text.canonicalization

      if entity.alternative_index is None or not entity.alternative_indices: entity # pylint: disable=pointless-statement continue - start = entity.range[const.START] - end = entity.range[const.END] + start = entity.range[const.EntityKeys.START] + end = entity.range[const.EntityKeys.END] transcript = transcripts[entity.alternative_index] canon = f"{transcript[:start]}{get_entity_type(entity)}{transcript[end:]}" for idx in entity.alternative_indices: @@ -295,9 +296,8 @@

      Source code for dialogy.plugins.text.canonicalization

      return canonicalized_transcripts
      -
      [docs] def utility(self, input: Input, output: Output) -> Any: - entities = output.entities - transcripts = input.transcripts +
      [docs] def utility(self, *args: Any) -> Any: + entities, transcripts = args return self.mask_transcript(entities, transcripts)
      [docs] def transform(self, training_data: pd.DataFrame) -> pd.DataFrame: @@ -308,7 +308,7 @@

      Source code for dialogy.plugins.text.canonicalization

      for i, row in tqdm(training_data.iterrows(), total=len(training_data)): try: - canonicalized_transcripts = self.mask_transcript( + canonicalized_transcripts = self.utility( row[self.entity_column], normalize(json.loads(row[self.input_column])), ) @@ -351,5 +351,7 @@

      Source code for dialogy.plugins.text.canonicalization

      + + \ No newline at end of file diff --git a/docs/_modules/dialogy/plugins/text/classification/xlmr.html b/docs/_modules/dialogy/plugins/text/classification/xlmr.html index b7663cd6..140a2938 100644 --- a/docs/_modules/dialogy/plugins/text/classification/xlmr.html +++ b/docs/_modules/dialogy/plugins/text/classification/xlmr.html @@ -5,7 +5,7 @@ - dialogy.plugins.text.classification.xlmr - dialogy 0.9.1 documentation + dialogy.plugins.text.classification.xlmr - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.1 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
    • Input
    • Output
    • Workflow
      • -
    • Base Plugin
      • +
    • Plugin
      • +
    • Intent
      • +
    • Entity
      • +
    • Slots

    Types

    Plugins

    Misc

      @@ -232,7 +234,7 @@

      Source code for dialogy.plugins.text.classification.xlmr

      from sklearn import preprocessing import dialogy.constants as const -from dialogy.base import Guard, Input, Output, Plugin +from dialogy.base.plugin import Plugin, PluginFn from dialogy.types import Intent from dialogy.utils import load_file, logger, save_file @@ -240,26 +242,21 @@

      Source code for dialogy.plugins.text.classification.xlmr

      [docs]class XLMRMultiClass(Plugin): """ This plugin provides a classifier based on `XLM-Roberta <https://arxiv.org/abs/1911.02116>`. - - .. _XLMRMultiClass: - The use_state flag in the XLMRMultiClass plugin is used to enable the use of state variable as the part of the text input. """ def __init__( self, model_dir: str, - dest: Optional[str] = None, - guards: Optional[List[Guard]] = None, + access: Optional[PluginFn] = None, + mutate: Optional[PluginFn] = None, debug: bool = False, threshold: float = 0.1, use_cuda: bool = False, score_round_off: int = 5, purpose: str = const.TRAIN, fallback_label: str = const.ERROR_LABEL, - use_state: bool = False, data_column: str = const.DATA, label_column: str = const.LABELS, - state_column: str = const.STATE, args_map: Optional[Dict[str, Any]] = None, skip_labels: Optional[List[str]] = None, kwargs: Optional[Dict[str, Any]] = None, @@ -273,7 +270,7 @@

      Source code for dialogy.plugins.text.classification.xlmr

      "Plugin requires simpletransformers -- https://simpletransformers.ai/docs/installation/" ) from error - super().__init__(dest=dest, guards=guards, debug=debug) + super().__init__(access, mutate, debug=debug) self.labelencoder = preprocessing.LabelEncoder() self.classifier = classifer self.model: Any = None @@ -281,9 +278,7 @@

      Source code for dialogy.plugins.text.classification.xlmr

      self.fallback_label = fallback_label self.data_column = data_column self.label_column = label_column - self.state_column = state_column self.use_cuda = use_cuda - self.use_state = use_state self.labelencoder_file_path = os.path.join( self.model_dir, const.LABELENCODER_FILE ) @@ -356,10 +351,9 @@

      Source code for dialogy.plugins.text.classification.xlmr

      def valid_labelencoder(self) -> bool: return hasattr(self.labelencoder, "classes_") -
      [docs] def inference(self, texts: Optional[List[str]]) -> List[Intent]: +
      [docs] def inference(self, texts: List[str]) -> List[Intent]: """ Predict the intent of a list of texts. - If the model has been trained using the state features, it expects the text to also be appended with the state token else the predictions would be spurious. :param texts: A list of strings, derived from ASR transcripts. :type texts: List[str] @@ -368,46 +362,37 @@

      Source code for dialogy.plugins.text.classification.xlmr

      :rtype: List[Intent] """ logger.debug(f"Classifier input:\n{texts}") - fallback_output = Intent(name=self.fallback_label, score=1.0).add_parser(self) - if not texts: - logger.error(f"texts passed to model {texts}!") - return [fallback_output] + fallback_output = Intent(name=self.fallback_label, score=1.0).add_parser( + self.__class__ + ) if self.model is None: logger.error(f"No model found for plugin {self.__class__.__name__}!") return [fallback_output] - if self.use_state: - for text in texts: - if "<st>" not in text: - logger.error( - f"use_state = True but {text} doesn't contain <st> </st> tags, \ - which means that training data is different from this data. This can lead to anomalous results" - ) - return [fallback_output] + if not self.valid_labelencoder: raise AttributeError( "Seems like you forgot to " f"save the {self.__class__.__name__} plugin." ) + if not texts: + return [fallback_output] + predictions, logits = self.model.predict(texts) if not predictions: return [fallback_output] - confidence_scores = [np.exp(logit) / sum(np.exp(logit)) for logit in logits] - intents_confidence_order = np.argsort(confidence_scores)[0][::-1] - predicted_intents = self.labelencoder.inverse_transform( - intents_confidence_order - ) - ordered_confidence_scores = [ - confidence_scores[0][idx] for idx in intents_confidence_order - ] + confidence_list = [max(np.exp(logit) / sum(np.exp(logit))) for logit in logits] + predicted_intents = self.labelencoder.inverse_transform(predictions) return [ Intent(name=intent, score=round(score, self.round)).add_parser( - self.__class__.__name__ + self.__class__ ) - for intent, score in zip(predicted_intents, ordered_confidence_scores) + if score > self.threshold + else fallback_output + for intent, score in zip(predicted_intents, confidence_list) ]
      [docs] def validate(self, training_data: pd.DataFrame) -> bool: @@ -415,7 +400,6 @@

      Source code for dialogy.plugins.text.classification.xlmr

      Validate the training data is in the appropriate format :param training_data: A pandas dataframe containing at least list of strings and corresponding labels. - Should also contain a state key if use_state = True while initializing object. :type training_data: pd.DataFrame :return: True if the dataframe is valid, False otherwise. :rtype: bool @@ -423,10 +407,8 @@

      Source code for dialogy.plugins.text.classification.xlmr

      if training_data.empty: logger.error("Training dataframe is empty.") return False - expected_columns = [self.data_column, self.label_column] - if self.use_state: - expected_columns.append(self.state_column) - for column in expected_columns: + + for column in [self.data_column, self.label_column]: if column not in training_data.columns: logger.warning(f"Column {column} not found in training data") return False @@ -437,7 +419,6 @@

      Source code for dialogy.plugins.text.classification.xlmr

      Train an intent-classifier on the provided training data. The training is skipped if the data-format is not valid. - While training with the use_state flag as true, make sure that the state column is the part of the training_data dataframe :param training_data: A pandas dataframe containing at least list of strings and corresponding labels. :type training_data: pd.DataFrame """ @@ -459,9 +440,6 @@

      Source code for dialogy.plugins.text.classification.xlmr

      training_data.loc[:, const.LABELS] = encoder.transform( training_data[const.LABELS] ) - # Add state as an additonal field to text - if self.use_state: - training_data[const.TEXT] += "<st> " + training_data["state"] + " </st>" training_data = training_data[[const.TEXT, const.LABELS]] self.init_model(len(encoder.classes_)) logger.debug( @@ -495,8 +473,8 @@

      Source code for dialogy.plugins.text.classification.xlmr

      self.labelencoder_file_path, mode="rb", loader=pickle.load )
      -
      [docs] def utility(self, input: Input, _: Output) -> List[Intent]: - return self.inference(input.clf_feature)
      +
      [docs] def utility(self, *args: Any) -> Any: + return self.inference(*args) # pylint: disable=no-value-for-parameter
      @@ -528,5 +506,7 @@

      Source code for dialogy.plugins.text.classification.xlmr

      + + \ No newline at end of file diff --git a/docs/_modules/dialogy/plugins/text/duckling_plugin.html b/docs/_modules/dialogy/plugins/text/duckling_plugin.html index 0dabe94c..e8af1bf2 100644 --- a/docs/_modules/dialogy/plugins/text/duckling_plugin.html +++ b/docs/_modules/dialogy/plugins/text/duckling_plugin.html @@ -5,7 +5,7 @@ - dialogy.plugins.text.duckling_plugin - dialogy 0.9.1 documentation + dialogy.plugins.text.duckling_plugin - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.1 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
    • Input
    • Output
    • Workflow
      • -
    • Base Plugin
      • +
    • Plugin
      • +
    • Intent
      • +
    • Entity
      • +
    • Slots

    Types

    Plugins

    Misc

      @@ -217,437 +219,56 @@

      Source code for dialogy.plugins.text.duckling_plugin

       """
-.. _DucklingPlugin:
-
-We use `Duckling <https://github.com/facebook/duckling>`_ for parsing values like: :code:`date`, :code:`time`,
-:code:`numbers`, :code:`currency` etc from natural language. The parser will expect Duckling to be running as an http service, and provide
-means to connect from the implementation here. Here's a big example for various duckling entities.
-
-Mother of all examples
-----------------------
-
-.. ipython::
-
-    In [1]: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.plugins import DucklingPlugin
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=[
-       ...:         "number",
-       ...:         "people",
-       ...:         "time",
-       ...:         "duration",
-       ...:         "intersect",
-       ...:         "amount-of-money",
-       ...:         "credit-card-number",
-       ...:     ],
-       ...:     # Duckling supports multiple dimensions, by specifying a list, we make sure
-       ...:     # we are searching for a match within the expected dimensions only.
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = [
-       ...:    "between today 7pm and tomorrow 9pm",
-       ...:    "can we get up at 6 am",
-       ...:    "we are 7 people",
-       ...:    "2 children 1 man 3 girls",
-       ...:    "can I come now?",
-       ...:    "can I come tomorrow",
-       ...:    "how about monday then",
-       ...:    "call me on 5th march",
-       ...:    "I want 4 pizzas",
-       ...:    "2 hours",
-       ...:    "can I pay $5 instead?",
-       ...:    "my credit card number is 4111111111111111",
-       ...:]
-
-    In [5]: %%time
-       ...: input_, output = workflow.run(Input(utterances=utterances))
-
-    In [6]: output
-
-
-Testing
--------
-
-1. Connect to the `Duckling API <https://github.com/facebook/duckling>`_ either via a docker container or a local setup.
-2. Boot an IPython session and setup an instance of :ref:`DucklingPlugin <DucklingPlugin>`.
+.. _duckling_plugin:
+
+This module exposes a parser for `Duckling <https://github.com/facebook/duckling>`_.
+
+`Duckling <https://github.com/facebook/duckling>`_ helps parsing values like: :code:`date`, :code:`time`,
+:code:`numbers`, :code:`currency` etc. The parser will expect Duckling to be running as an http service, and provide
+means to connect from the implementation here.
 
 .. code-block:: python
     :linenos:
+    :emphasize-lines: 4, 30
 
-    from pprint import pprint
     from dialogy.workflow import Workflow
-    from dialogy.base import Input
-    from dialogy.plugins import DucklingPlugin
+
+    def update_workflow(workflow, entities):
+        workflow.input["duckling_entities"] = entities
 
     duckling_plugin = DucklingPlugin(
-        dest="output.entities",
-        dimensions=["people", "time"],
+        access=lambda workflow: (workflow.input["sentence"], workflow.input["reftime"], workflow.input["locale"])
+        # the access method guides the plugin to data within a workflow.
+        # in this case, the `input` property of a workflow is expected
+        # to be a `dict` with a key named "sentence". Check line 30.
+        mutate=update_workflow,
+        # the mutate method guides the plugin to place the processed output
+        # somewhere on the workflow.
+        dimensions=["people"],
+        # Duckling supports multiple dimensions, by specifying a list, we make sure
+        # we are searching for a match within the expected dimensions only.
         locale="en_IN",
+        # Duckling supports a set of locales, we need to provide this info to
+        # get language specific matches.
         timezone="Asia/Kolkata",
-        origin/master
-    )
-
-    entities = duckling_plugin.parse(["We are 2 children coming tomorrow at 5 pm"])
-    pprint(entities)
-    # [PeopleEntity(
-    #     body='2 children', 
-    #     type='value', 
-    #     parsers=['DucklingPlugin'], 
-    #     score=1.0, 
-    #     alternative_index=0,
-    #     alternative_indices=[0, 0], 
-    #     latent=False, 
-    #     value=2, 
-    #     origin='value', 
-    #     unit='child',
-    #     entity_type='people')
-    # ,TimeEntity(
-    #     body='tomorrow at 5 pm', 
-    #     type='value', 
-    #     parsers=['DucklingPlugin'], 
-    #     score=1.0, 
-    #     alternative_index=0, 
-    #     alternative_indices=[0],
-    #     latent=False, 
-    #     value='2022-02-11T17:00:00.000+05:30', origin='value', entity_type='datetime', grain='hour')]
-
-    # To convert these to dicts:
-
-    entity_dicts = [entity.json() for entity in entities]
-    pprint(entity_dicts)
-    # [{'range': {'start': 7, 'end': 17},
-    #     'body': '2 children',
-    #     'type': 'value',
-    #     'parsers': ['DucklingPlugin'],
-    #     'score': 1.0,
-    #     'alternative_index': 0,
-    #     'value': 2,
-    #     'unit': 'child',
-    #     'entity_type': 'people'},
-    # {'range': {'start': 38, 'end': 54},
-    #     'body': 'tomorrow at 5 pm',
-    #     'type': 'value',
-    #     'parsers': ['DucklingPlugin'],
-    #     'score': 1.0,
-    #     'alternative_index': 0,
-    #     'value': '2022-02-11T17:00:00.000+05:30',
-    #     'entity_type': 'datetime',
-    #     'grain': 'hour'
-    # }]
-
-Guards
-------
-
-Like all plugins, the Duckling plugin can be :ref:`guarded <Guards>` by conditions that prevent its execution.
-You may need this if you want to save on latency and are sure of not expecting entities in those turns. We will use
-the :ref:`mother of all examples <Mother of all examples>` from above to make the point. We will use the :code:`current_state` to tell the plugin
-to guard itself in the *"SMALL_TALK"* state.
-
-.. ipython::
-
-    In [1]: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.plugins import DucklingPlugin
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=[
-       ...:         "number",
-       ...:         "people",
-       ...:         "date",
-       ...:         "time",
-       ...:         "duration",
-       ...:         "intersect",
-       ...:         "amount-of-money",
-       ...:         "credit-card-number",
-       ...:     ],
-       ...:     guards=[lambda i, o: i.current_state == "SMALL_TALK"]
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = [
-       ...:    "between today 7pm and tomorrow 9pm",
-       ...:    "can we get up at 6 am",
-       ...:    "we are 7 people",
-       ...:    "2 children 1 man 3 girls",
-       ...:    "can I come now?",
-       ...:    "can I come tomorrow",
-       ...:    "how about monday then",
-       ...:    "call me on 5th march",
-       ...:    "I want 4 pizzas",
-       ...:    "2 hours",
-       ...:    "can I pay $5 instead?",
-       ...:    "my credit card number is 4111111111111111",
-       ...:]
-
-    In [5]: %%time
-       ...: input_, output = workflow.run(Input(utterances=utterances, current_state="SMALL_TALK"))
-
-    In [6]: output
-
-You can notice that the time drops from **ms** to **us** and we produce no entities.
-
-Filtering Date and Time
------------------------
-
-There are cases where we run into ambiguity in entity resolution. A fun case happens usually with
-the Hindi language when someone says *"कल"*. The problem is, *"कल"* could be *"tommorrow"* or *"yesterday"*.
-
-Problem
-^^^^^^^
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time"],
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = ["कल"]
-
-    In [5]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-
-    In [5]: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-    In [6]: output
-
-See the two entities? 
-
-- We froze our reference time for 1st Jan 2022 at 12:00:00. 
-- We get 31st December 2021
-- and 2nd Jan 2022.
-
-This could be decoded using additional context. Let's assume it is a given that dates can only be in future. Say
-a flight-booking case.
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time"],
-       ...:     datetime_filters=DucklingPlugin.FUTURE
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = ["कल"]
-
-    In [5]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-
-    In [5]: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-    In [6]: output
-
-We can use :code:`DucklingPlugin.PAST` for inverse. 
-
-What if the context was within the utterance? Say for payment verification cases, we receive these utterances:
-
-- *"कल ही कर दी थी"* -- *"I have already paid yesterday"*
-- *"कल कर दूंगा"* -- *"I will pay tomorrow"*
-
-We can't filter out either future or past values. In these cases, the `intent <Intents>` carries the context
-for resolving the entity value. We will look into :ref:`Temporal Intents <Temporal Intents>` for more details.
-
-
-Temporal Intents
------------------
-
-Filtering
-^^^^^^^^^
-
-There are cases where intents carry a notion of time, repeating an example from above:
-
-#. *"कल ही कर दी थी"* -- *"I have already paid yesterday"*
-#. *"कल कर दूंगा"* -- *"I will pay tomorrow"*
-
-While we can't decode the entity value for "कल" in the above utterances, the correct intent prediction is:
-
-#. :code:`already_paid`
-#. :code:`pay_later`
-
-If we could use this information, we could resolve the entity value for "कल". We will simulate the output
-to contain appropriate intents using the :code:`set` method of :ref:`workflow <Workflow>`.
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.types import Intent
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time"],
-       ...:     temporal_intents={
-       ...:         "already_paid": DucklingPlugin.FUTURE,
-       ...:         "pay_later": DucklingPlugin.PAST
-       ...:     }
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = ["कल ही कर दी थी"]
-       ...: workflow.set("output.intents", [Intent(name="already_paid", score=1.0)])
-
-    In [5]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-       ...: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-    In [6]: output
-
-    In [7]: utterances = ["कल ही कर दी थी"]
-       ...: workflow.set("output.intents", [Intent(name="pay_later", score=1.0)])
-       ...: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-    In [6]: output
-
-Casting
-^^^^^^^
-
-We can also use temporal intents to cast certain entities. Machines can act on absolute time values but 
-natural language often has relative units like *"for 2 hours"*. These produce a duration entity.
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["duration", "time"],
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: utterances = ["for 2h"]
-
-    In [5]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-
-    In [5]: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-    In [6]: output
-
-We understand the duration but we could create an absolute time value using the reference time, but we don't know if duration 
-should be added or removed.
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.types import Intent
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time", "duration"],
-       ...:     timezone="Asia/Kolkata",
-       ...:     temporal_intents={
-       ...:         "already_paid": DucklingPlugin.FUTURE,
-       ...:         "pay_later": DucklingPlugin.PAST
-       ...:     }
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-       ...: utterances = ["for 2h"]
-
-    In [5]: workflow.set("output.intents", [Intent(name="already_paid", score=1.0)])
-       ...: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-    In [6]: output
-
-    In [7]: workflow.set("output.intents", [Intent(name="pay_later", score=1.0)])
-       ...: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-    In [8]: output
-
-In case we need to always cast duration as a future or a past value we can do:
-
-
-.. ipython::
-
-    In [1]: import pytz
-       ...: from datetime import datetime
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.base import Input
-       ...: from dialogy.types import Intent
-       ...: from dialogy.plugins import DucklingPlugin
-       ...: from dialogy.utils import dt2timestamp
-
-    In [2]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time", "duration"],
-       ...:     timezone="Asia/Kolkata",
-       ...:     temporal_intents={
-       ...:         "__any__": DucklingPlugin.FUTURE,
-       ...:     }
-       ...: )
-
-    In [3]: workflow = Workflow([duckling_plugin])
-
-    In [4]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-       ...: reference_time = dt2timestamp(reference_dt)
-       ...: utterances = ["for 2h"]
-
-    In [5]: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-    In [6]: output
-
-    In [7]: duckling_plugin = DucklingPlugin(
-       ...:     dest="output.entities",
-       ...:     dimensions=["time", "duration"],
-       ...:     timezone="Asia/Kolkata",
-       ...:     temporal_intents={
-       ...:         "__any__": DucklingPlugin.PAST,
-       ...:     }
-       ...: )
-
-    In [8]: workflow = Workflow([duckling_plugin])
-       ...: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-       ...: output
+        # Date/Time related entities make this field imperative.
+    )() # ⬅️ the instance is also getting called.
+    # Notice that we instantiate the class and also
+    # invoke __call__ method of the instance.
+    # The returned value from the instance is the plugin.
+    #
+    # This is a standard pattern across all plugins.
+
+    workflow = Workflow(preprocessors=[duckling_plugin])
+    workflow.run(input_={"sentence": "there are 7 people"})
+    # Once an input is run through the workflow as on line 30,
+    # we can expect the plugin mutation to create an artifact
+    # as per line 4.
 
 """
 import json
 import operator
 import traceback
-from concurrent import futures
 from datetime import datetime
 from pprint import pformat
 from typing import Any, Callable, Dict, List, Optional, Union
@@ -660,15 +281,22 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      from tqdm import tqdm
 
 from dialogy import constants as const
-from dialogy.base import Guard, Input, Output, Plugin
 from dialogy.base.entity_extractor import EntityScoringMixin
-from dialogy.types import BaseEntity, Intent
-from dialogy.types.entity.deserialize import deserialize_duckling_entity
+from dialogy.base.plugin import Plugin, PluginFn
+from dialogy.constants import EntityKeys
+from dialogy.types.entity import BaseEntity, dimension_entity_map
 from dialogy.utils import dt2timestamp, lang_detect_from_text, logger
 
 
 
      [docs]class DucklingPlugin(EntityScoringMixin, Plugin):
     """
+    A :ref:`Plugin<plugin>` for extracting entities using `Duckling <https://github.com/facebook/duckling>`_. Once
+    instantiated, a :code:`duckling_parser` object will interface to an http server, running `Duckling
+    <https://github.com/facebook/duckling>`_.
+
+    This object when used as a plugin, transforms the :code:`List[Dict[str, Any]]` returned from the API to a
+    :ref:`BaseEntity<base_entity>` or one of its subclasses.
+
     :param dimensions: `Dimensions <https://github.com/facebook/duckling#supported-dimensions>`_. Of the listed
     dimensions, we support:
 
@@ -704,30 +332,18 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
          :type url: Optional[str]
     """
 
-    FUTURE = const.FUTURE
-    """
-    Select time entity values with scores greater than or equal to the reference time.
-    """
-
-    PAST = const.PAST
-    """
-    Select time entity values with scores lesser than or equal to the reference time.
-    """
-
-    __DATETIME_OPERATION_ALIAS = {FUTURE: operator.ge, PAST: operator.le}
-
     def __init__(
         self,
         dimensions: List[str],
-        timezone: str = "UTC",
+        timezone: str,
         timeout: float = 0.5,
         url: str = "http://0.0.0.0:8000/parse",
         locale: str = "en_IN",
-        temporal_intents: Optional[Dict[str, str]] = None,
-        dest: Optional[str] = None,
-        guards: Optional[List[Guard]] = None,
         datetime_filters: Optional[str] = None,
         threshold: Optional[float] = None,
+        access: Optional[PluginFn] = None,
+        mutate: Optional[PluginFn] = None,
+        entity_map: Optional[Dict[str, Any]] = None,
         activate_latent_entities: Union[Callable[..., bool], bool] = False,
         reference_time_column: str = const.REFERENCE_TIME,
         input_column: str = const.ALTERNATIVES,
@@ -739,8 +355,8 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
              constructor
         """
         super().__init__(
-            dest=dest,
-            guards=guards,
+            access=access,
+            mutate=mutate,
             debug=debug,
             input_column=input_column,
             output_column=output_column,
@@ -752,22 +368,19 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      self.timeout = timeout
         self.threshold = threshold
         self.url = url
-        self.temporal_intents = temporal_intents or {}
         self.reference_time: Optional[int] = None
         self.reference_time_column = reference_time_column
         self.datetime_filters = datetime_filters
         self.activate_latent_entities = activate_latent_entities
-        self.session = requests.Session()
-        self.session.mount(
-            "http://",
-            requests.adapters.HTTPAdapter(
-                max_retries=1, pool_maxsize=10, pool_block=True
-            ),
-        )
         self.headers: Dict[str, str] = {
             "Content-Type": "application/x-www-form-urlencoded; charset=UTF-8"
         }
 
+        if isinstance(entity_map, dict):
+            self.dimension_entity_map = {**dimension_entity_map, **entity_map}
+        else:
+            self.dimension_entity_map = dimension_entity_map
+
     def __set_timezone(self) -> Optional[BaseTzInfo]:
         """
         Set timezone as BaseTzInfo from compatible timezone string.
@@ -795,7 +408,7 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      text: str,
         reference_time: Optional[int] = None,
         locale: str = "en_IN",
-        use_latent: Union[Callable[..., bool], bool] = False,
+        use_latent: bool = False,
     ) -> Dict[str, Any]:
         """
         create request body for entity parsing
@@ -835,16 +448,6 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      return payload
 
-
      [docs]    def get_operator(self, filter_type: Any) -> Any:
-        try:
-            return getattr(operator, filter_type)
-        except (AttributeError, TypeError) as exception:
-            logger.debug(traceback.format_exc())
-            raise ValueError(
-                f"Expected datetime_filters to be one of {self.FUTURE}, {self.PAST} "
-                "or a valid comparison operator here: https://docs.python.org/3/library/operator.html"
-            ) from exception
      
-
 
      [docs]    def select_datetime(
         self, entities: List[BaseEntity], filter_type: Any
     ) -> List[BaseEntity]:
@@ -864,10 +467,17 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      " a valid comparison operator here: https://docs.python.org/3/library/operator.html"
             )
 
-        if filter_type in self.__DATETIME_OPERATION_ALIAS:
-            operation = self.__DATETIME_OPERATION_ALIAS[filter_type]
+        if filter_type in self.DATETIME_OPERATION_ALIAS:
+            operation = self.DATETIME_OPERATION_ALIAS[filter_type]
         else:
-            operation = self.get_operator(filter_type)
+            try:
+                operation = getattr(operator, filter_type)
+            except (AttributeError, TypeError) as exception:
+                logger.debug(traceback.format_exc())
+                raise ValueError(
+                    f"Expected datetime_filters to be one of {self.FUTURE}, {self.PAST} "
+                    "or a valid comparison operator here: https://docs.python.org/3/library/operator.html"
+                ) from exception
 
         time_entities, other_entities = py_.partition(
             entities, lambda entity: entity.dim == const.TIME
@@ -903,11 +513,7 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      return super().apply_filters(entities)
      
 
     def _reshape(
-        self,
-        entities_json: List[Dict[str, Any]],
-        alternative_index: int = 0,
-        reference_time: Optional[int] = None,
-        duration_cast_operator: Optional[str] = None,
+        self, entities_json: List[Dict[str, Any]], alternative_index: int = 0
     ) -> List[BaseEntity]:
         """
         Create a list of :ref:`BaseEntity <base_entity>` objects from a list of entity dicts.
@@ -919,27 +525,55 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
              :return: A list of objects subclassed from :ref:`BaseEntity <base_entity>`
         :rtype: Optional[List[BaseEntity]]
         """
-        deserialized_entities = []
-        for entity_json in entities_json:
-            entity = deserialize_duckling_entity(
-                entity_json,
-                alternative_index,
-                reference_time=reference_time,
-                timezone=self.timezone,
-                duration_cast_operator=duration_cast_operator,
-            )
-            entity.add_parser(self)
-            deserialized_entities.append(entity)
-        return deserialized_entities
+        entity_object_list: List[BaseEntity] = []
+
+        try:
+            # For each entity dict:
+            for entity in entities_json:
+                if entity[EntityKeys.VALUE][EntityKeys.TYPE] in [
+                    EntityKeys.VALUE,
+                    EntityKeys.DURATION,
+                    EntityKeys.INTERVAL,
+                ]:
+                    # We can auto convert dict forms of duckling entities to dialogy entity classes only if they are
+                    # known in advance. We currently support only the types in the condition above.
+                    if entity[EntityKeys.VALUE][EntityKeys.TYPE] == EntityKeys.INTERVAL:
+                        # Duckling entities with interval type have a different structure for value(s).
+                        # They have a need to express units in "from", "to" format.
+                        cls = self.dimension_entity_map[entity[EntityKeys.DIM]][EntityKeys.INTERVAL]  # type: ignore
+                    else:
+                        cls = self.dimension_entity_map[entity[EntityKeys.DIM]][EntityKeys.VALUE]  # type: ignore
+                    # The most appropriate class is picked for making an object from the dict.
+                    duckling_entity: BaseEntity = cls.from_dict(entity)
+                    duckling_entity.add_parser(self)
+                    # Depending on the type of entity, the value is searched and filled.
+                    duckling_entity.set_value()
+                    duckling_entity.alternative_index = alternative_index
+                    # Collect the entity object in a list.
+                    entity_object_list.append(duckling_entity)
+                else:
+                    # Raised only if an unsupported `dimension` is used.
+                    raise NotImplementedError(
+                        f"Entities with value.type {entity['value']['type']} are"
+                        " not implemented. Report this"
+                        " issue here: https://github.com/Vernacular-ai/dialogy/issues"
+                    )
+        except KeyError as key_error:
+            # Being vary of structural changes in the API or entity dicts.
+            # Under normal circumstances this error shouldn't be raised.
+            logger.debug(traceback.format_exc())
+            raise KeyError(
+                f"Missing key {key_error} in entity {entity}."
+            ) from key_error
+        return entity_object_list
 
     def _get_entities(
         self,
         text: str,
         locale: str = "en_IN",
         reference_time: Optional[int] = None,
-        use_latent: Union[Callable[..., bool], bool] = False,
-        sort_idx: int = 0,
-    ) -> Dict[str, Any]:
+        use_latent: bool = False,
+    ) -> List[Dict[str, Any]]:
         """
         Get entities from duckling-server.
 
@@ -953,7 +587,7 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
                                      to parse the values into usable dates/times, defaults to None
         :type reference_time: Optional[int], optional
         :raises ValueError: Duckling API call failure leading to no json response.
-        :return: Duckling entities as :code:`dicts`.
+        :return: Duckling entities as python :code:`dicts`.
         :rtype: List[Dict[str, Any]]
         """
         body = self.__create_req_body(
@@ -961,22 +595,22 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      )
 
         try:
-            response = self.session.post(
+            response = requests.post(
                 self.url, data=body, headers=self.headers, timeout=self.timeout
             )
 
             if response.status_code == 200:
                 # The API call was successful, expect the following to contain entities.
                 # A list of dicts or an empty list.
-                return {const.IDX: sort_idx, const.VALUE: response.json()}
+                return response.json()
         except requests.exceptions.Timeout as timeout_exception:
-            logger.error(f"Duckling timed out: {timeout_exception}")  # pragma: no cover
-            logger.error(pformat(body))  # pragma: no cover
-            return {const.IDX: sort_idx, const.VALUE: []}  # pragma: no cover
+            logger.error(f"Duckling timed out: {timeout_exception}")
+            logger.error(pformat(body))
+            return []
         except requests.exceptions.ConnectionError as connection_error:
             logger.error(f"Duckling server is turned off?: {connection_error}")
             logger.error(pformat(body))
-            raise requests.exceptions.ConnectionError from connection_error
+            return []
 
         # Control flow reaching here would mean the API call wasn't successful.
         # To prevent rest of the things from crashing, we will raise an exception.
@@ -984,76 +618,19 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      f"Duckling API call failed | [{response.status_code}]: {response.text}"
         )
 
-    def _get_entities_concurrent(
-        self,
-        texts: List[str],
-        locale: str = "en_IN",
-        reference_time: Optional[int] = None,
-        use_latent: Union[Callable[..., bool], bool] = False,
-    ) -> List[List[Dict[str, Any]]]:
+
      [docs]    def utility(self, *args: Any) -> List[BaseEntity]:
         """
-        Make multiple-parallel API calls to duckling-server .
+        Produces Duckling entities, runs with a :ref:`Workflow's run<workflow_run>` method.
 
-        :param texts: A list of strings to be parsed concurrently via duckling.
-        :type texts: List[str]
-        :param locale: Defined in __init__, defaults to "en_IN"
-        :type locale: str, optional
-        :param reference_time: Cases where relative units of time are mentioned,
-                                like "today", "now", etc. We need to know the current time
-                                to parse the values into usable dates/times, defaults to None
-        :type reference_time: Optional[int], optional
-        :param use_latent: True returns latent entities, defaults to False
-        :type use_latent: bool, optional
-        :return: Duckling entities as :code:`dicts`.
-        :rtype: List[List[Dict[str, Any]]]
+        :param args: Expects a tuple of :code:`Tuple[natural language for parsing entities, reference time in milliseconds, locale]`
+        :type args: Tuple(Union[str, List[str]], int, str)
+        :return: A list of duckling entities.
+        :rtype: List[BaseEntity]
         """
-        workers = min(10, max(len(texts), 1))
-        with futures.ThreadPoolExecutor(max_workers=workers) as executor:
-            futures_ = [
-                executor.submit(
-                    self._get_entities,
-                    text,
-                    locale,
-                    reference_time=reference_time,
-                    use_latent=use_latent,
-                    sort_idx=i,
-                )
-                for i, text in enumerate(texts)
-            ]
-        entities_list = [future.result() for future in futures.as_completed(futures_)]
-        return [
-            entities[const.VALUE]
-            for entities in sorted(
-                entities_list, key=lambda entities: entities[const.IDX]
-            )
-        ]
+        list_of_entities: List[List[Dict[str, Any]]] = []
+        shaped_entities: List[List[BaseEntity]] = []
 
-
      [docs]    def apply_entity_classes(
-        self,
-        list_of_entities: List[List[Dict[str, Any]]],
-        reference_time: Optional[int] = None,
-        timezone: str = "UTC",
-        duration_cast_operator: Optional[str] = None,
-    ) -> List[BaseEntity]:
-        shaped_entities = []
-        for (alternative_index, entities) in enumerate(list_of_entities):
-            shaped_entities.append(
-                self._reshape(
-                    entities,
-                    alternative_index=alternative_index,
-                    reference_time=reference_time,
-                    duration_cast_operator=duration_cast_operator,
-                )
-            )
-        return py_.flatten(shaped_entities)
      
-
-
      [docs]    def validate(
-        self, input_: Union[str, List[str]], reference_time: Optional[int]
-    ) -> "DucklingPlugin":
-        input_is_str = isinstance(input_, str)
-        inputs_are_list_of_strings = isinstance(input_, list) and all(
-            isinstance(text, str) for text in input_
-        )
+        input_, reference_time, locale, use_latent = args
         if not isinstance(reference_time, int) and self.datetime_filters:
             raise TypeError(
                 "Duckling requires reference_time to be a unix timestamp (int) but"
@@ -1061,75 +638,44 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      "https://stackoverflow.com/questions/20822821/what-is-a-unix-timestamp-and-why-use-it\n"
             )
 
-        if not input_is_str and not inputs_are_list_of_strings:
-            raise TypeError(f"Expected {input_} to be a List[str] or str.")
-        return self
      
+        self.reference_time = reference_time
+        input_size = 1
 
-
      [docs]    def parse(
-        self,
-        transcripts: Union[str, List[str]],
-        locale: Optional[str] = None,
-        reference_time: Optional[int] = None,
-        use_latent: Union[Callable[..., bool], bool] = False,
-        intents: Optional[List[Intent]] = None,
-    ) -> List[BaseEntity]:
-        list_of_entities: List[List[Dict[str, Any]]] = []
-        entities: List[BaseEntity] = []
-        duration_cast_operator = None
-        intent = None
+        try:
+            if isinstance(input_, str):
+                list_of_entities.append(
+                    self._get_entities(
+                        input_,
+                        locale,
+                        reference_time=reference_time,
+                        use_latent=use_latent,
+                    )
+                )
+            elif isinstance(input_, list) and all(
+                isinstance(text, str) for text in input_
+            ):
+                input_size = len(input_)
+                for text in input_:
+                    entities = self._get_entities(
+                        text,
+                        locale,
+                        reference_time=reference_time,
+                        use_latent=use_latent,
+                    )
+                    list_of_entities.append(entities)
+            else:
+                raise TypeError(f"Expected {input_} to be a List[str] or str.")
 
-        if intents:
-            intent, *_ = intents
+            for (alternative_index, entities) in enumerate(list_of_entities):
+                shaped_entities.append(self._reshape(entities, alternative_index))
 
-        if isinstance(intent, Intent) and intent.name in self.temporal_intents:
-            duration_cast_operator = self.datetime_filters = self.temporal_intents.get(
-                intent.name
+            shaped_entities_flattened = py_.flatten(shaped_entities)
+            aggregate_entities = self.entity_consensus(
+                shaped_entities_flattened, input_size
             )
-
-        if const.ANY in self.temporal_intents:
-            duration_cast_operator = self.temporal_intents.get(const.ANY)
-
-        locale = locale or self.locale
-        self.reference_time = reference_time = reference_time or self.reference_time
-        self.validate(transcripts, reference_time)
-        if isinstance(transcripts, str):
-            transcripts = [transcripts]  # pragma: no cover
-
-        list_of_entities = self._get_entities_concurrent(
-            transcripts,
-            locale=locale,
-            reference_time=reference_time,
-            use_latent=use_latent,
-        )
-        entities = self.apply_entity_classes(
-            list_of_entities,
-            reference_time,
-            duration_cast_operator=duration_cast_operator,
-        )
-        entities = self.entity_consensus(entities, len(transcripts))
-        return self.apply_filters(entities)
      
-
-
      [docs]    def utility(self, input: Input, output: Output) -> List[BaseEntity]:
-        """
-        Produces Duckling entities, runs with a :ref:`Workflow's run<workflow_run>` method.
-
-        :param argbrightons: Expects a tuple of :code:`Tuple[natural language for parsing entities, reference time in milliseconds, locale]`
-        :type args: Tuple(Union[str, List[str]], int, str)
-        :return: A list of duckling entities.
-        :rtype: List[BaseEntity]
-        """
-        transcripts = input.transcripts
-        self.reference_time = input.reference_time
-        self.locale = input.locale or self.locale
-        use_latent = input.latent_entities
-
-        return self.parse(
-            transcripts,
-            locale=self.locale,
-            reference_time=self.reference_time,
-            use_latent=use_latent,
-            intents=output.intents,
-        )
      
+            return self.apply_filters(aggregate_entities)
+        except ValueError as value_error:
+            raise ValueError(str(value_error)) from value_error
      
 
 
      [docs]    def transform(self, training_data: pd.DataFrame) -> pd.DataFrame:
         """
@@ -1162,11 +708,11 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      f"{reference_time=} should be isoformat date or unix timestamp integer."
                 )
             transcripts = self.make_transform_values(row[self.input_column])
-            entities = self.parse(
+            entities = self.utility(
                 transcripts,
+                reference_time,
                 lang_detect_from_text(self.input_column),
-                reference_time=reference_time,
-                use_latent=self.activate_latent_entities,
+                self.activate_latent_entities,
             )
             if row[self.output_column] is None or pd.isnull(row[self.output_column]):
                 training_data.at[i, self.output_column] = entities
@@ -1207,5 +753,7 @@ 
      Source code for dialogy.plugins.text.duckling_plugin
      
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/plugins/text/list_entity_plugin.html b/docs/_modules/dialogy/plugins/text/list_entity_plugin.html
index e7a045c1..56ffe48a 100644
--- a/docs/_modules/dialogy/plugins/text/list_entity_plugin.html
+++ b/docs/_modules/dialogy/plugins/text/list_entity_plugin.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.plugins.text.list_entity_plugin - dialogy 0.9.1 documentation
+        dialogy.plugins.text.list_entity_plugin - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.1 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.1 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -160,11 +160,13 @@
 
    • Input
      • 
 
    • Output
      • 
 
    • Workflow
      • 
-
    • Base Plugin
      • 
+
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

    Misc

      @@ -217,77 +219,27 @@

      Source code for dialogy.plugins.text.list_entity_plugin

       """
-.. _ListEntityPlugin:
-
-Regex Search
--------------
-
-We have often seen certain keywords that gain significance in an SLU project. These keywords are 
-easy to extract via patterns and are often used to create entities. The :ref:`ListEntityPlugin<ListEntityPlugin>`
-helps in this task, it requires a pattern-map, we call it :code:`candidates`.
-
-
-.. ipython::
-
-    In [1]: from dialogy.workflow import Workflow
-       ...: from dialogy.plugins import ListEntityPlugin
-       ...: from dialogy.base import Input
-
-    In [2]: candidates = {
-       ...:      "colours": {
-       ...:          "red": ["red", "crimson", "ruby", "raspberry"],
-       ...:          "blue": ["blue", "azure", "indigo", "navy"],
-       ...:          "green": ["green", "emerald", "jade", "teal"],
-       ...:      }, 
-       ...:    }
+.. _list_entity_plugin:
 
-    In [2]: list_entity_plugin = ListEntityPlugin(
-       ...:   candidates=candidates,
-       ...:   style="regex",
-       ...:   dest="output.entities"
-       ...: )
-       ...: workflow = Workflow([list_entity_plugin])
-       ...: _, output = workflow.run(Input(utterances="I want an emerald green shirt"))
+Module needs refactor. We are currently keeping all strategies bundled as methods as opposed to SearchStrategyClasses.
 
-    In [3]: output
-
-Spacy NER
------------
-
-We also allow using `spacy's NER <https://spacy.io/usage/spacy-101#annotations-ner>`_. This has to be passed within the
-:code:`spacy_nlp` attribute.
-
-.. ipython::
-
-    In [1]: import spacy
-       ...: from dialogy.workflow import Workflow
-       ...: from dialogy.plugins import ListEntityPlugin
-       ...: from dialogy.base import Input
-
-    In [2]: nlp = spacy.load("en_core_web_sm")
-
-    In [3]: list_entity_plugin = ListEntityPlugin(
-       ...:   spacy_nlp=nlp,
-       ...:   style="spacy",
-       ...:   dest="output.entities"
-       ...: )
-       ...: workflow = Workflow([list_entity_plugin])
-       ...: _, output = workflow.run(Input(utterances="Need a place to stay in New Delhi."))
-
-    In [3]: output
+Within dialogy, we extract entities using Duckling, Pattern lists and Spacy. We can ship individual plugins but at the
+same time, the difference is just configuration of each of these tools/services. There is another difference of
+intermediate structure that the DucklingPlugin expects. We need to prevent the impact of the structure from affecting
+all other entities. So that their :code:`from_dict(...)` methods are pristine and involve no shape hacking.
 """
 import re
 from pprint import pformat
 from typing import Any, Dict, List, Optional, Tuple
 
 import pandas as pd
-from loguru import logger
 from tqdm import tqdm
 
 from dialogy import constants as const
-from dialogy.base import Guard, Input, Output, Plugin
 from dialogy.base.entity_extractor import EntityScoringMixin
+from dialogy.base.plugin import Plugin, PluginFn
 from dialogy.types import BaseEntity, KeywordEntity
+from dialogy.utils import logger
 
 Text = str
 Label = str
@@ -298,25 +250,28 @@ 
      Source code for dialogy.plugins.text.list_entity_plugin
      [docs]class ListEntityPlugin(EntityScoringMixin, Plugin):
     """
-     A :ref:`Plugin<AbstractPlugin>` for extracting entities using spacy or a list of regex patterns.
-
-     .. attention:
+     A :ref:`Plugin<plugin>` for extracting entities using spacy or a list of regex patterns.
 
-        This class will undergo a series of refactoring changes. FWIW, :ref:`ListSearchPlugin<ListSearchPlugin>`
-        is more more performant in terms of entity capture rates but not as responsive. :code:`ListEntityPlugin`
-        is fast. So make choices with bearing this in mind.
+     .. note: This class will undergo a series of refactoring changes. The final form will accommodate Duckling, Spacy
+        and regex based entity parsers.
 
     :param style: One of ["regex", "spacy"]
     :type style: Optional[str]
     :param candidates: Required if style is "regex", this is a :code:`dict` that shows a mapping of entity
         values and their patterns.
     :type candidates: Optional[Dict[str, List[str]]]
-    :param spacy_nlp: Required if style is "spacy", requires is a
+    :param spacy_nlp: Required if style is "spacy", this is a
         `spacy model <https://spacy.io/usage/spacy-101#annotations-ner>`_.
     :type spacy_nlp: Any
     :param labels: Required if style is "spacy". If there is a need to extract only a few labels from all the other
         `available labels <https://github.com/explosion/spaCy/issues/441#issuecomment-311804705>`_.
     :type labels: Optional[List[str]]
+    :param access: A plugin io utility that allows access to transcripts
+        :code:`List[str]` within a :ref:`Workflow <workflow>`.
+    :type access: Optional[PluginFn]
+    :param mutate: A plugin io utility that allows insertion of :code:`List[BaseEntity]` within a
+        :ref:`Workflow <workflow>`.
+    :type mutate: Optional[PluginFn]
     :param debug: A flag to set debugging on the plugin methods
     :type debug: bool
     """
@@ -326,10 +281,10 @@ 
      Source code for dialogy.plugins.text.list_entity_plugin
      style: Optional[str] = None,
         candidates: Optional[Dict[str, Dict[str, List[Any]]]] = None,
         spacy_nlp: Any = None,
-        dest: Optional[str] = None,
-        guards: Optional[List[Guard]] = None,
         labels: Optional[List[str]] = None,
         threshold: Optional[float] = None,
+        access: Optional[PluginFn] = None,
+        mutate: Optional[PluginFn] = None,
         input_column: str = const.ALTERNATIVES,
         output_column: Optional[str] = None,
         use_transform: bool = True,
@@ -337,8 +292,8 @@ 
      Source code for dialogy.plugins.text.list_entity_plugin
      debug: bool = False,
     ):
         super().__init__(
-            dest=dest,
-            guards=guards,
+            access=access,
+            mutate=mutate,
             debug=debug,
             input_column=input_column,
             output_column=output_column,
@@ -452,25 +407,25 @@ 
      Source code for dialogy.plugins.text.list_entity_plugin
      for i, matches_on_transcript in enumerate(matches_on_transcripts):
             for text, label, value, span in matches_on_transcript:
                 entity_dict = {
-                    const.RANGE: {
-                        const.START: span[0],
-                        const.END: span[1],
-                    },
-                    const.BODY: text,
-                    const.DIM: label,
-                    const.SCORE: 0,
-                    const.ALTERNATIVE_INDEX: i,
-                    const.LATENT: False,
+                    "start": span[0],
+                    "end": span[1],
+                    "body": text,
+                    "dim": label,
+                    "parsers": [self.__class__.__name__],
+                    "score": 0,
+                    "alternative_index": i,
+                    "latent": False,
                     "__group": f"{label}_{text}",
-                    const.TYPE: label,
-                    const.ENTITY_TYPE: label,
-                    const.VALUE: value,
-                    const.VALUES: [{const.VALUE: value}],
+                    "type": label,
+                    "entity_type": label,
+                    "value": {
+                        "values": [{"value": value}],
+                    },
                 }
 
                 del entity_dict["__group"]
                 entity_ = KeywordEntity.from_dict(entity_dict)
-                entity_.add_parser(self)
+                entity_.add_parser(self).set_value()
                 entities.append(entity_)
 
         logger.debug("Parsed entities")
@@ -479,9 +434,8 @@ 
      Source code for dialogy.plugins.text.list_entity_plugin
      aggregated_entities = self.entity_consensus(entities, len(transcripts))
         return self.apply_filters(aggregated_entities)
      
 
-
      [docs]    def utility(self, input: Input, _: Output) -> Any:
-        transcripts = input.transcripts
-        return self.get_entities(transcripts)  # pylint: disable=no-value-for-parameter
      
+
      [docs]    def utility(self, *args: Any) -> Any:
+        return self.get_entities(*args)  # pylint: disable=no-value-for-parameter

    Types

    Plugins

    Misc

      @@ -217,33 +219,34 @@

      Source code for dialogy.plugins.text.merge_asr_output

       """
-We use classifiers for prediction of :ref:`intents<Intent>`, utterances from an ASR are prominent features
-for the task. We featurize the utterances by concatenation of all transcripts.
-
-.. ipython::
-
-    In [1]: from dialogy.workflow import Workflow
-       ...: from dialogy.plugins import MergeASROutputPlugin
-       ...: from dialogy.base import Input
-
-    In [2]: merge_asr_output_plugin = MergeASROutputPlugin(dest="input.clf_feature")
-       ...: workflow = Workflow([merge_asr_output_plugin])
-
-    In [3]: input_, _ = workflow.run(Input(utterances=["we will come by 7 pm", " will come by 7 pm"]))
-
-    In [4]: input_
+We expect ASR transcripts to be the most common inputs to the system. Since there is a chance that we can improve the
+performance by asking the ASR to provide more information in the form of multiple alternatives. The problem though, is
+they can't be used as it is. Models that can operate on text data, need them to be vectorized. These vectors have
+semantic significance, where different text inputs imply different vectors. If their semantic differences are modeled
+well, we can expect these vectors to be similar or dissimilar just like the text that was used to create them.
+
+The problem with transcripts is, there **will** be noise in the transcripts, since we asked for :code:`N` transcripts,
+by definition, an ASR will provide N unique transcripts. The noise may in fact be inconsequential for downstream NLP
+tasks, but we have noticed that there are certain inputs owing to audio artifacts or state of LMs tend to have a great
+impact on the transcription.
+
+This module will ship a simple concatenation strategy to address this problem. The NLP model should be expected to
+handle documents (:code:`List[str]`) instead of single text. The plugin then will be able to produce a :code:`str` by
+simply concatenation of each transcript.
+
+This may not be very helpful, there is a :ref:`VotePlugin <vote_plugin>` under development so use it with caution. It may
+help with precision at the cost of recall.
 """
 import json
 import traceback
 from typing import Any, List, Optional
 
 import pandas as pd
-from loguru import logger
 from tqdm import tqdm
 
 import dialogy.constants as const
-from dialogy.base import Guard, Input, Output, Plugin
-from dialogy.utils import normalize
+from dialogy.base.plugin import Plugin, PluginFn
+from dialogy.utils import logger, normalize
 
 
 # == merge_asr_output ==
@@ -264,6 +267,13 @@ 
      Source code for dialogy.plugins.text.merge_asr_output
          The normalization is done by :ref:`normalize<normalize>`
 
+    .. ipython:: python
+
+        from dialogy.plugins.text.merge_asr_output import merge_asr_output
+
+        utterances = ["This is a sentence", "That is a sentence"]
+        merge_asr_output(utterances)
+
     :param utterances: A structure representing ASR output. We support only:
 
         1. :code:`List[str]`
@@ -292,9 +302,9 @@ 
      Source code for dialogy.plugins.text.merge_asr_output
      [docs]class MergeASROutputPlugin(Plugin):
     """
-    .. _MergeASROutputPlugin:
+    .. _merge_asr_output_plugin:
 
-    Working details are covered in :ref:`MergeASROutputPlugin<MergeASROutputPlugin>`.
+    Working details are covered in :ref:`merge_asr_output <merge_asr_output>`.
 
     :param Plugin: [description]
     :type Plugin: [type]
@@ -302,24 +312,24 @@ 
      Source code for dialogy.plugins.text.merge_asr_output
      def __init__(
         self,
+        access: Optional[PluginFn],
+        mutate: Optional[PluginFn],
         input_column: str = const.ALTERNATIVES,
         output_column: Optional[str] = None,
         use_transform: bool = False,
-        dest: Optional[str] = None,
-        guards: Optional[List[Guard]] = None,
         debug: bool = False,
     ) -> None:
         super().__init__(
-            dest=dest,
-            guards=guards,
+            access=access,
+            mutate=mutate,
+            debug=debug,
             input_column=input_column,
             output_column=output_column,
             use_transform=use_transform,
-            debug=debug,
         )
 
-
      [docs]    def utility(self, input: Input, _: Output) -> Any:
-        return merge_asr_output(input.utterances)
      
+
      [docs]    def utility(self, *args: Any) -> Any:
+        return merge_asr_output(*args)
      
 
 
      [docs]    def transform(self, training_data: pd.DataFrame) -> pd.DataFrame:
         if not self.use_transform:
@@ -378,5 +388,7 @@ 
      Source code for dialogy.plugins.text.merge_asr_output
      
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/plugins/text/slot_filler/rule_slot_filler.html b/docs/_modules/dialogy/plugins/text/slot_filler/rule_slot_filler.html
index 540dfcbb..e9b5123c 100644
--- a/docs/_modules/dialogy/plugins/text/slot_filler/rule_slot_filler.html
+++ b/docs/_modules/dialogy/plugins/text/slot_filler/rule_slot_filler.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.plugins.text.slot_filler.rule_slot_filler - dialogy 0.9.2 documentation
+        dialogy.plugins.text.slot_filler.rule_slot_filler - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.2 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.2 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -160,11 +160,13 @@
 
    • Input
      • 
 
    • Output
      • 
 
    • Workflow
      • 
-
    • Base Plugin
      • 
+
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

    Misc

      @@ -217,58 +219,118 @@

      Source code for dialogy.plugins.text.slot_filler.rule_slot_filler

       """
-Slot Filling
-============
-
-While we describe :ref:`slots <Slot>` in the linked doc. The slot filling is part of an :ref:`intent's<Intent>`
-role. This plugin orchestrates the execution of slot filling via methods available on intents.
+.. _rule_slot_filler:
 
-#. We eagerly apply slot placeholders using the :ref:`apply<ApplySlot>` method. 
-#. Once entities are found, we check if the slots can fill them using the :ref:`fill<FillSlot>` method.
-#. If no slots were filled, we remove the placeholders using :ref:`cleanup<CleanupSlot>` method.
+Module provides access to a rule-based :ref:`slot filler<slot_filler>`.
 """
-from typing import List, Optional
+from typing import Any, List, Optional
 
-from dialogy.base import Guard, Input, Output, Plugin
-from dialogy.types import BaseEntity
+from dialogy.base.plugin import Plugin
+from dialogy.types.entity import BaseEntity
 from dialogy.types.intent import Intent
+from dialogy.types.plugin import PluginFn
 from dialogy.types.slots import Rule
+from dialogy.utils.logger import logger
 
 
 
      [docs]class RuleBasedSlotFillerPlugin(Plugin):
     """
-    A utility :ref:`plugin <AbstractPlugin>` for
+    A utility :ref:`plugin <plugin>` for
     `slot filling <https://nlpprogress.com/english/intent_detection_slot_filling.html>`._
     An :ref:`Intent <intent>` may have a few slots that need to be filled.
 
     This plugin can assist filling pertaining to certain intent:entity:slot-name association rules.
 
-    :ref:`Schema for rules<ApplySlot>`
+    Schema for rules looks like:
+
+    .. code-block:: json
+
+        {
+            "intent_name": {
+                "slot_name":"entity_type"
+            }
+        }
+
+    This can be represented in a yaml format like so:
+
+    .. code-block:: yaml
+
+        slots:
+            faqs:
+                action_slot: actions
+            report:
+                time_slot:
+                - time
+                - number
 
     Let's run through a practical example. We will create a workflow and preset the output to have expected intent and
     entities.
 
-    .. _RuleBasedSlotFillerPlugin:
+    .. ipython:: python
+
+        import yaml
+        from pprint import pprint
+        from dialogy.workflow import Workflow
+        from dialogy.plugins import RuleBasedSlotFillerPlugin
+        from dialogy.types.intent import Intent
+        from dialogy.types.entity import PeopleEntity, NumericalEntity
+
+        # ---------------------------------------------------------------------------------------------------
+        # Create rules
+        rule_string = \"""
+        faqs:
+            number_slot: number
+        count_people:
+            number_slot:
+            - people
+            - number
+        \"""
+        slot_rules = yaml.safe_load(rule_string)
+        # ---------------------------------------------------------------------------------------------------
+
+        # ---------------------------------------------------------------------------------------------------
+        # Setting up the plugin
+        slot_filler = RuleBasedSlotFillerPlugin(rules=slot_rules, access=lambda w: w.output, debug=True)
+        # ---------------------------------------------------------------------------------------------------
+
+        # ---------------------------------------------------------------------------------------------------
+        # Creating synthetic values.
+        # normally you'd get these by running the workflow with models.
+        intent = Intent(name="count_people", score=1)
+        number_entity = NumericalEntity(type="number", range={"start": 0, "end": 7}, body='2 people', latent=False, values=[{"values":2}], dim="number")
+        people_entity = PeopleEntity(type="people", range={"start": 0, "end": 7}, body='2 people', latent=False, values=[{"values":2}], dim="people")
+        entities = [people_entity, number_entity]
+        # ---------------------------------------------------------------------------------------------------
+
+        # ---------------------------------------------------------------------------------------------------
+        # setting up the workflow to use slot_filler
+        workflow = Workflow([slot_filler], debug=True)
+        # ---------------------------------------------------------------------------------------------------
+
+        # If you notice the instantiation of RuleBasedSlotFillerPlugin,
+        # you will notice the `access` method expects a Tuple of two elements.
+        workflow.output = (intent, entities)
+
+        workflow.run(input_="2 am")
 
-    :param rules: A mapping that defines relationship between an intent, its slots and the entities that fill them.
-    :type rules: Rule
-    :param fill_multiple: More than one item be allowed to fill a single slot.
-    :type fill_multiple: bool
-    :param access: Signature for workflow access is :code:`access(workflow) -> (Intent, List[BaseEntity])`
-    :type access: Optional[PluginFn]
-    :param mutate: Not needed, pass :code:`None`.
-    :type mutate: Optional[PluginFn]
+    :param rules: A mapping that defines relationship between an intent, its slots and the entities that fill them.
+    :type rules: Rule
+    :param fill_multiple: More than one item be allowed to fill a single slot.
+    :type fill_multiple: bool
+    :param access: Signature for workflow access is :code:`access(workflow) -> (Intent, List[BaseEntity])`
+    :type access: Optional[PluginFn]
+    :param mutate: Not needed, pass :code:`None`.
+    :type mutate: Optional[PluginFn]
 
-    Irrespective of the entities that are found, only the listed type in the slot shall be present in `values`.
-    """
+    Irrespective of the entities that are found, only the listed type in the slot shall be present in `values`.
+    """
 
     def __init__(
         self,
         rules: Rule,
-        dest: Optional[str] = None,
-        guards: Optional[List[Guard]] = None,
-        replace_output: bool = True,
-        fill_multiple: bool = True,
+        fill_multiple: bool = False,
+        access: Optional[PluginFn] = None,
+        mutate: Optional[PluginFn] = None,
         debug: bool = False,
     ) -> None:
         """
@@ -283,9 +345,7 @@ 
      Source code for dialogy.plugins.text.slot_filler.rule_slot_filler
      # ```
         # rules = {"intent": {"slot_name": "entity_type"}}
         # ```
-        super().__init__(
-            dest=dest, guards=guards, debug=debug, replace_output=replace_output
-        )
+        super().__init__(access=access, mutate=mutate, debug=debug)
         self.rules: Rule = rules or {}
 
         # fill_multiple
@@ -293,22 +353,24 @@ 
      Source code for dialogy.plugins.text.slot_filler.rule_slot_filler
      # same entity type within a slot.
         self.fill_multiple = fill_multiple
 
-
      [docs]    def fill(self, intents: List[Intent], entities: List[BaseEntity]) -> List[Intent]:
+
      [docs]    def fill(self, intents: List[Intent], entities: List[BaseEntity]) -> None:
         if not intents:
-            return intents
+            return
 
-        intent, *rest = intents
+        if not (isinstance(intents, list) and isinstance(intents[0], Intent)):
+            return
 
+        intent = intents[0]
         intent.apply(self.rules)
 
         for entity in entities:
             intent.fill_slot(entity, fill_multiple=self.fill_multiple)
 
         intent.cleanup()
-        return [intent, *rest]
      
+        logger.debug(f"intent after slot-filling: {intent}")
      
 
-
      [docs]    def utility(self, _: Input, output: Output) -> List[Intent]:
-        return self.fill(output.intents, output.entities)
      
+
      [docs]    def utility(self, *args: Any) -> Any:
+        return self.fill(*args)  # pylint: disable=no-value-for-parameter
    @@ -340,5 +402,7 @@

    Source code for dialogy.plugins.text.slot_filler.rule_slot_filler

    + + \ No newline at end of file diff --git a/docs/_modules/dialogy/types/entity/base_entity.html b/docs/_modules/dialogy/types/entity/base_entity.html index d11e0d11..bcf46bc1 100644 --- a/docs/_modules/dialogy/types/entity/base_entity.html +++ b/docs/_modules/dialogy/types/entity/base_entity.html @@ -5,7 +5,7 @@ - dialogy.types.entity.base_entity - dialogy 0.9.3 documentation + dialogy.types.entity.base_entity - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -147,7 +147,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -161,11 +161,12 @@
  • Output
  • Workflow
  • Plugin
    • +
  • Intent
    • +
  • Entity
    • +
  • Slots

    • Types

    Plugins

      @@ -219,90 +219,24 @@

      Source code for dialogy.types.entity.base_entity

       """
-.. _BaseEntity:
-
+.. _base_entity:
 Module provides access to a base class to create other entities.
 
+These methods of this class that are supposed to be overridden:
+- validate
+- get_value
 
-.. mermaid::
-
-    classDiagram
-        direction LR
-        BaseEntity --|> NumericalEntity
-        BaseEntity --|> CreditCardNumberEntity
-        BaseEntity --|> CurrencyEntity    
-        BaseEntity --|> TimeEntity
-        BaseEntity --|> KeywordEntity
-        BaseEntity --|> DurationEntity
-        BaseEntity --|> PeopleEntity
-
-        TimeEntity --|> TimeIntervalEntity
-
-        NumericalEntity ..> TimeEntity: as_time
-        DurationEntity ..> TimeEntity: as_time
-
-        class BaseEntity {
-            +Dict[str, int] range
-            +str body
-            +str dim
-            +int alternative_index
-            +str type
-            +str entity_type
-            +float score
-            +Any value
-            +List~Any~ values
-            +List~str~ parsers
-            +bool latent
-            +Any get_value()
-            +Dict[str, Any] json()
-            $BaseEntity from_dict(d)
-        }
-
-        class CurrencyEntity {
-            +str unit
-            +Any get_value()
-            +CurrencyEntity from_duckling(d)
-        }
-
-        class CreditCardNumberEntity {
-            +str issuer
-            +from_duckling(d: dict, idx: int)
-            +CreditCardNumberEntity from_duckling(d)
-        }
-
-        class KeywordEntity {
-        }
-
-        class DurationEntity {
-            +str unit
-            +dict normalized
-            +DurationEntity from_duckling(d)
-        }
-
-        class NumericalEntity {
-            +NumericalEntity from_duckling(d)
-        }
-
-        class PeopleEntity {
-            +str unit
-            +PeopleEntity from_duckling(d)
-        }
-
-        class TimeEntity {
-            +str grain
-            +datetime get_value()
-            +TimeEntity from_duckling(d)
-        }
+Import classes:
 
+- BaseEntity
 """
-from __future__ import annotations
-
 import copy
 from typing import Any, Dict, List, Optional, Union
 
 import attr
 
 from dialogy import constants as const
+from dialogy.utils import traverse_dict, validate_type
 
 
 # = BaseEntity =
@@ -318,72 +252,140 @@ 
      Source code for dialogy.types.entity.base_entity
      
     # **range**
     #
     # is the character range in the alternative where the entity is parsed.
-    range: Dict[str, int] = attr.ib(repr=False, order=False, kw_only=True)
+    range = attr.ib(type=Dict[str, int], repr=False)
+
+    # **type**
+    #
+    # is same as dimension or `dim` for now. We may deprecate `dim` and retain only `type`.
+    type = attr.ib(type=str, validator=attr.validators.instance_of(str))
 
     # **body**
     #
     # is the string from which the entity is extracted.
-    body: str = attr.ib(validator=attr.validators.instance_of(str), order=False)
+    body = attr.ib(type=str, validator=attr.validators.instance_of(str))
 
     # **dim**
     #
     # is influenced from Duckling's convention of categorization.
-    dim: Optional[str] = attr.ib(default=None, repr=False, order=False)
-
-    # **type**
-    #
-    # is same as dimension or `dim` for now. We may deprecate `dim` and retain only `type`.
-    type: str = attr.ib(default="value", order=False, kw_only=True)
+    dim = attr.ib(type=Optional[str], default=None, repr=False)
 
     # **parsers**
     #
     # gives the trail of all the functions that have changed this entity in the sequence of application.
-    parsers: List[str] = attr.ib(
+    parsers = attr.ib(
+        type=List[str],
         default=attr.Factory(list),
         validator=attr.validators.instance_of(list),
-        order=False,
     )
 
     # **score**
     #
     # is the confidence that the range is the entity.
-    score: Optional[float] = attr.ib(default=None)
+    score = attr.ib(type=Optional[float], default=None)
+
+    # **slot_names**
+    #
+    # Entities have awareness of the slots they should fill.
+    slot_names = attr.ib(type=List[str], default=attr.Factory(list), repr=False)
 
     # **alternative_index**
     #
     # is the index of transcript within the ASR output: `List[Utterances]`
     # from which this entity was picked up. This may be None.
-    alternative_index: Optional[int] = attr.ib(default=None, order=False)
-    alternative_indices: Optional[List[int]] = attr.ib(default=None, order=False)
+    alternative_index = attr.ib(type=Optional[int], default=None)
+    alternative_indices = attr.ib(type=Optional[List[int]], default=None)
 
     # **latent**
     #
     # Duckling influenced attribute, tells if there is less evidence for an entity if latent is True.
-    latent: bool = attr.ib(default=False, order=False)
+    latent = attr.ib(type=bool, default=False)
 
     # **values**
     #
     # The parsed value of an entity resides within this attribute.
-    values: List[Dict[str, Any]] = attr.ib(
+    values = attr.ib(
+        type=List[Dict[str, Any]],
         default=attr.Factory(list),
-        validator=attr.validators.optional(attr.validators.instance_of(list)),  # type: ignore
+        validator=attr.validators.instance_of(List),
         repr=False,
-        order=False,
     )
 
     # **values**
     #
     # A single value interpretation from values.
-    value: Any = attr.ib(default=None, order=False)
+    value: Any = attr.ib(default=None)
 
     # **entity_type**
-    entity_type: Optional[str] = attr.ib(default=None, repr=True, order=False)
+    #
+    # Mirrors type, to be deprecated.
+    entity_type: Optional[str] = attr.ib(default=None, repr=False)
+
+    __properties_map = const.BASE_ENTITY_PROPS
 
     def __attrs_post_init__(self) -> None:
-        if self.values and not self.value:
-            self.value = self.values[0][const.VALUE]
-        elif self.value and not self.values:
-            self.values = [{const.VALUE: self.value}]
+        self.entity_type = self.type
+
+
      [docs]    @classmethod
+    def validate(cls, dict_: Dict[str, Any]) -> None:
+        """
+        Check attributes of instance match expected types.
+
+        :param dict_: A `Dict` where each key is an attribute of the instance and value is the expected type.
+        :type dict_: Dict[str, Any]
+        :return: None
+        :rtype: None
+        """
+        for prop, prop_type in cls.__properties_map:
+            validate_type(traverse_dict(dict_, prop), prop_type)
      
+
+
      [docs]    @classmethod
+    def reshape(cls, dict_: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        :type dict_: Dict[str, Any]
+        """
+        match_start = dict_[const.EntityKeys.START]
+        match_end = dict_[const.EntityKeys.END]
+
+        dict_[const.EntityKeys.RANGE] = {
+            const.EntityKeys.START: match_start,
+            const.EntityKeys.END: match_end,
+        }
+        # ['body', 'start', 'value', 'end', 'dim', 'latent']
+
+        # **type** of an entity is same as its **dimension**.
+        dict_[const.EntityKeys.TYPE] = dict_[const.EntityKeys.DIM]
+
+        # This piece is a preparation for multiple entity values.
+        # So, even though we are confident of the value found, we are still keeping the
+        # structure.
+        if const.EntityKeys.VALUES in dict_[const.EntityKeys.VALUE]:
+            dict_[const.EntityKeys.VALUES] = dict_[const.EntityKeys.VALUE][
+                const.EntityKeys.VALUES
+            ]
+        else:
+            dict_[const.EntityKeys.VALUES] = [dict_[const.EntityKeys.VALUE]]
+
+        del dict_[const.EntityKeys.START]
+        del dict_[const.EntityKeys.END]
+        del dict_[const.EntityKeys.VALUE]
+        return dict_
      
+
+
      [docs]    @classmethod
+    def from_dict(cls, dict_: Dict[str, Any]) -> "BaseEntity":
+        """
+        Create an instance of a given class `cls` from a `dict` that complies
+        with attributes of `cls` through its keys and values.
+        Compliance is verified using the `__validate` method. It is expected that each subclass
+        will implement their own flavor of `__validate` to check their respective inputs.
+
+        :param dict_: A dict that provides all the attributes necessary for instantiating this class
+        :type dict_: Dict[str, Any]
+        :return: Instance of class
+        :rtype: BaseEntity
+        """
+        dict_ = cls.reshape(dict_)
+        cls.validate(dict_)
+        return cls(**dict_)
      
 
 
      [docs]    def add_parser(self, plugin: Union[Any, str]) -> "BaseEntity":
         """
@@ -397,10 +399,11 @@ 
      Source code for dialogy.types.entity.base_entity
      
         :return: Calling instance with modifications to :code:`parsers` attribute.
         :rtype: BaseEntity
         """
-        self.parsers.append(str(plugin))
+        plugin_name = plugin if isinstance(plugin, str) else plugin.__class__.__name__
+        self.parsers.append(plugin_name)
         return self
      
 
-
      [docs]    def get_value(self) -> Any:
+
      [docs]    def get_value(self, reference: Any = None) -> Any:
         """
         Get value of an entity.
 
@@ -416,12 +419,17 @@ 
      Source code for dialogy.types.entity.base_entity
      
         """
         key = "value"
         error_message = f'entity value should be in values[0]["{key}"]'
-        try:
-            return self.values[0][key]
-        except IndexError as index_error:
-            raise IndexError(error_message) from index_error
      
-
-
      [docs]    def copy(self) -> BaseEntity:
+        if not reference:
+            try:
+                return self.values[0][key]
+            except IndexError as index_error:
+                raise IndexError(error_message) from index_error
+            except KeyError as key_error:
+                raise KeyError(error_message) from key_error
+        else:
+            return reference.get(const.VALUE)
      
+
+
      [docs]    def copy(self) -> "BaseEntity":
         """
         Create a deep copy of the instance and return.
 
@@ -455,16 +463,21 @@ 
      Source code for dialogy.types.entity.base_entity
      
             skip_ = [name for name in const.SKIP_ENTITY_ATTRS if name not in add]
         return attr.asdict(self, filter=lambda attr, _: attr.name not in skip_)
      
 
-
      [docs]    @classmethod
-    def from_dict(
-        cls, dict_: Dict[str, Any], reference: Optional[BaseEntity] = None
-    ) -> BaseEntity:
-        if reference:
-            if const.VALUES in dict_ or const.VALUE in dict_:
-                object.__setattr__(reference, const.VALUES, None)
-                object.__setattr__(reference, const.VALUE, None)
-            return attr.evolve(reference, **dict_)
-        return cls(**dict_)
      
+
      [docs]    def set_value(self, value: Any = None) -> "BaseEntity":
+        """
+        Set values and value attribute.
+
+        :param value: The parsed value of an entity token.
+        :type value: Any
+        :return: None
+        :rtype: None
+        """
+        if value is None and isinstance(self.values, list) and len(self.values) > 0:
+            self.value = self.values[0][const.VALUE]
+        else:
+            self.values = [{const.VALUE: value}]
+            self.value = value
+        return self
      
 
 
 # = entity_synthesis =
diff --git a/docs/_modules/dialogy/types/intent.html b/docs/_modules/dialogy/types/intent.html
index d3004983..bae2ffc4 100644
--- a/docs/_modules/dialogy/types/intent.html
+++ b/docs/_modules/dialogy/types/intent.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.types.intent - dialogy 0.9.1 documentation
+        dialogy.types.intent - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.1 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.1 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -160,11 +160,13 @@
 
    • Input
      • 
 
    • Output
      • 
 
    • Workflow
      • 
-
    • Base Plugin
      • 
+
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

    Misc

      @@ -217,124 +219,68 @@

      Source code for dialogy.types.intent

       """
-.. _Intent:
-
-Intent Classification
-======================
-
-Automatic categorization of texts into known classes. When solving SLU problems
-the objective is to answer *What was meant?* instead of *"What was said?"*. 
-
-For an example utterance *"I lost my wallet"* a bank's support executives would 
-understand that they must follow the operating procedure for blocking all transaction 
-accounts. We want to train machines to have the ability to interpret these cues.
+.. _intent:
+Type definition for `Intent`.
 
-We call the classes of *meaning* as :code:`Intent`.
+Import classes:
 
-We can obtain intents from :ref:`XLMRMultiClass<XLMRMultiClass>` and :ref:`MLPMultiClass<MLPMultiClass>` plugin.
+    - Intent
 """
-from __future__ import annotations
 
 from typing import Any, Dict, List, Optional
 
-import attr
-
-from dialogy.types import BaseEntity
+from dialogy.types.entity import BaseEntity
 from dialogy.types.slots import Rule, Slot
 from dialogy.utils.logger import logger
 
 
-
      [docs]@attr.s
-class Intent:
+
      [docs]class Intent:
     """
     An instance of this class contains the name of the action associated with a body of text.
     """
 
-    # The name of the intent to be used.
-    name: str = attr.ib(kw_only=True, order=False)
-    """
-    The description of an intent separated by '_'
-    """
+    def __init__(
+        self,
+        name: str,
+        score: float,
+        parsers: Optional[List[str]] = None,
+        alternative_index: Optional[int] = 0,
+        slots: Optional[Dict[str, Slot]] = None,
+    ) -> None:
+        """
 
-    # The confidence of this intent being present in the utterance.
-    score: float = attr.ib(kw_only=True, default=0.0, order=True)
-    """
-    A positive real number that represents confidence of this intent
-    being the meaning of an utterance. Higher is better.
-    
-    Should be a value between 0 and 1.
-    """
+        :param name: The name of the intent. A class label
+        :type name: str
+        :param score: The confidence score from a model.
+        :type score: float
+        :param parsers: Items that alter the attributes.
+        :type parsers: Optional[List[str]]
+        :param alternative_index: Out of a list of transcripts, this points at the index which lead to prediction.
+        :type alternative_index: Optional[int]
+        :param slots: A map of slot names and :ref:`Slot<slot>`
+        :type slots: Optional[Dict[str, Slot]]
+        """
+        # The name of the intent to be used.
+        self.name = name
 
-    # In case of an ASR, `alternative_index` points at one of the nth
-    # alternatives that help in predictions.
-    alternative_index: Optional[int] = attr.ib(kw_only=True, default=None, order=False)
-    """
-    Tells us the index of transcript that yeilds this intent. Since our featurizer
-    concatenates all transcripts, we currently don't have a use for it.
+        # The confidence of this intent being present in the utterance.
+        self.score = score
 
-    If we were to predict the intent for each transcript separately and vote, in those
-    cases it would be helpful to log this property.
-    """
+        # Trail of functions that modify the attributes of an instance.
+        self.parsers = parsers or []
 
-    # Trail of functions that modify the attributes of an instance.
-    parsers: List[str] = attr.ib(kw_only=True, factory=list, order=False, repr=False)
-    """
-    Contains a list of plugins that have created, updated this Intent.
-    """
+        # In case of an ASR, `alternative_index` points at one of the nth
+        # alternatives that help in predictions.
+        self.alternative_index = alternative_index
 
-    # Container for holding `List[BaseEntity]`.
-    slots: Dict[str, Slot] = attr.ib(
-        kw_only=True, factory=dict, order=False, repr=False
-    )
-    """
-    Placeholders for :ref:`entities <BaseEntity>` that are relevant to this intent. 
-    More details are available in the doc for :ref:`slots <Slot>`.
-    """
+        # Container for holding `List[BaseEntity]`.
+        self.slots: Dict[str, Slot] = slots or {}
 
-
      [docs]    def apply(self, rules: Rule) -> Intent:
+
      [docs]    def apply(self, rules: Rule) -> "Intent":
         """
-        Create slots using rules.
-
-        .. _ApplySlot:
+        Create slots using :ref:`rules`.
 
         An intent can hold different entities within associated slot-types.
-        We parse rules that describe this mapping and create placeholders within an intent for
-        holding entities found.
-
-        An example rule would look like:
-
-        .. code:: python
-
-            rules = {
-                "wake_up_alarm": {          # name of the intent.
-                    "datetime_slot": [      # "datetime_slot" is the only slot associated with this intent.
-                        "date",             # Only "date", "time", "datetime" entities can fill the "datetime_slot"
-                        "time",
-                        "datetime"
-                    ]
-                },
-                "ticket_booking": {         # name of the intent.
-                    "origin": [             # A slot that will fill only location type entities.
-                        "location"          # The entity types that can be filled in the slot.
-                    ],
-                    "destination": [        # Another slot that will fill only location type entities.
-                        "location"
-                    ],
-                    "start_date": [
-                        "date"
-                    ],
-                    "end_date": [
-                        "date"
-                    ],
-                    "passengers": [
-                        "people",
-                        "number"
-                    ],
-                    "budget": [
-                        "amount-of-money"
-                    ]
-                }
-            }
 
         :param rules: A configuration for slot names and entity types associated with this instance of Intent.
         :type rules: Rule
@@ -365,7 +311,7 @@ 
      Source code for dialogy.types.intent
       
         return self
      
 
-
      [docs]    def add_parser(self, plugin: Any) -> Intent:
+
      [docs]    def add_parser(self, plugin: Any) -> "Intent":
         """
         Update parsers with the plugin name
 
@@ -381,12 +327,10 @@ 
      Source code for dialogy.types.intent
               self.parsers.append(plugin_name)
         return self
      
 
-
      [docs]    def fill_slot(self, entity: BaseEntity, fill_multiple: bool = False) -> Intent:
+
      [docs]    def fill_slot(self, entity: BaseEntity, fill_multiple: bool = False) -> "Intent":
         """
         Update :code:`slots[slot_type].values` with a single entity.
 
-        .. _FillSlot:
-
         We will explore the possibility of sharing/understanding the meaning of multiple entities
         in the same slot type.
 
@@ -410,11 +354,10 @@ 
      Source code for dialogy.types.intent
                       f"slot type: {slot.types}",
             )
             logger.debug(
-                f"entity type: {entity.entity_type}",
+                f"entity type: {entity.type}",
             )
-            if entity.entity_type in slot.types:
+            if entity.type in slot.types:
                 if fill_multiple:
-                    logger.debug(f"filling {entity} into {self.name}.")
                     self.slots[slot_name].add(entity)
                     return self
 
@@ -434,8 +377,6 @@ 
      Source code for dialogy.types.intent
       
      [docs]    def cleanup(self) -> None:
         """
         Remove slots that were not filled.
-
-        .. _CleanupSlot:
         """
         slot_names = list(self.slots.keys())
         for slot_name in slot_names:
@@ -455,11 +396,14 @@ 
      Source code for dialogy.types.intent
               """
         return {
             "name": self.name,
-            "alternative_index": self.alternative_index,
             "score": self.score,
+            "alternative_index": self.alternative_index,
+            "slots": [slot.json() for slot in self.slots.values()],
             "parsers": self.parsers,
-            "slots": {slot_name: slot.json() for slot_name, slot in self.slots.items()},
-        }
      
+        }
      
+
+    def __repr__(self) -> str:
+        return f"Intent(name={self.name}, score={self.score}, slots={self.slots})"
    @@ -491,5 +435,7 @@

    Source code for dialogy.types.intent

         
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/types/slots.html b/docs/_modules/dialogy/types/slots.html
index e42c6bd8..ee337cb7 100644
--- a/docs/_modules/dialogy/types/slots.html
+++ b/docs/_modules/dialogy/types/slots.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.types.slots - dialogy 0.9.1 documentation
+        dialogy.types.slots - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
    -
    dialogy 0.9.1 documentation
     +
    dialogy 0.9.4 documentation
    @@ -147,7 +147,7 @@
    - dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
  • Input
  • Output
  • Workflow
    • -
  • Base Plugin
    • +
  • Plugin
    • +
  • Intent
    • +
  • Entity
    • +
  • Slots

    • Types

    Plugins

    Misc

      @@ -217,23 +219,23 @@

      Source code for dialogy.types.slots

       """
-Type definition for Slots. Read :ref:`Intent<Intent>` and :ref:`Slot Filling<Slot Filling>` for details.
-"""
-from __future__ import annotations
+.. _slot:
 
-from typing import Any, Dict, List
+Type definition for Slots.
 
-import attr
+Import classes:
 
-from dialogy.types import BaseEntity
+    - Slot
+"""
 
+from typing import Any, Dict, List
 
-
      [docs]@attr.s
-class Slot:
-    """
-    Slot Type
+import dialogy.constants as const
+from dialogy.types.entity import BaseEntity
 
-    .. _Slot:
+
+
      [docs]class Slot:
+    """Slot Type
 
     Keys are:
     - `name` of the slot
@@ -241,18 +243,19 @@ 
      Source code for dialogy.types.slots
           - `values` list of entities extracted
     """
 
-    name: str = attr.ib(kw_only=True, order=True)
-    types: List[str] = attr.ib(kw_only=True, factory=list, order=False)
-    values: List[BaseEntity] = attr.ib(kw_only=True, factory=list, order=False)
+    def __init__(self, name: str, types: List[str], values: List[BaseEntity]) -> None:
+        self.name = name
+        self.types = types
+        self.values = values
 
-
      [docs]    def add(self, entity: BaseEntity) -> Slot:
+
      [docs]    def add(self, entity: BaseEntity) -> "Slot":
         """
         Insert the `BaseEntity` within the current `Slot` instance.
         """
         self.values.append(entity)
         return self
      
 
-
      [docs]    def clear(self) -> Slot:
+
      [docs]    def clear(self) -> "Slot":
         """
         Remove all `BaseEntity` within the current `Slot` instance.
         """
@@ -266,11 +269,16 @@ 
      Source code for dialogy.types.slots
               Returns:
             Dict[str, Any]
         """
-        return {
+        entities_json = [entity.json() for entity in self.values]
+        slot_json = {
             "name": self.name,
-            "types": self.types,
-            "values": [entity.json() for entity in self.values],
-        }
      
+            "type": self.types,
+            const.EntityKeys.VALUES: entities_json,
+        }
+        return slot_json
      
+
+    def __repr__(self) -> str:
+        return f"Slot(name={self.name}, types={self.types}, values={self.values})"
      
 
 
 Rule = Dict[str, Dict[str, Any]]
@@ -305,5 +313,7 @@ 
      Source code for dialogy.types.slots
           
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/utils/file_handler.html b/docs/_modules/dialogy/utils/file_handler.html
index 056dd003..10a7dfe5 100644
--- a/docs/_modules/dialogy/utils/file_handler.html
+++ b/docs/_modules/dialogy/utils/file_handler.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.utils.file_handler - dialogy 0.9.1 documentation
+        dialogy.utils.file_handler - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
       
     
      
     
      
-      
      dialogy 0.9.1 documentation
      
+      
      dialogy 0.9.4 documentation
      
     
      
     
      
       
      
@@ -147,7 +147,7 @@
       
      
   
   
-  dialogy 0.9.1 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -160,11 +160,13 @@
 
    • Input
      • 
 
    • Output
      • 
 
    • Workflow
      • 
-
    • Base Plugin
      • 
+
    • Plugin
      • 
+
    • Intent
      • 
+
    • Entity
      • 
+
    • Slots
      •

    Types

    Plugins

    Misc

    Types

    Plugins

    Misc

      @@ -223,6 +225,7 @@

      Source code for dialogy.utils.misc

           - dict_traversal
     - validate_type
 """
+from datetime import datetime
 from functools import reduce
 from typing import Any, Dict, List, Tuple, Union
 
@@ -296,6 +299,18 @@ 
      Source code for dialogy.utils.misc
           """
     if not isinstance(obj, obj_type):
         raise TypeError(f"{obj} should be a {obj_type}")
      
+
+
+
      [docs]def dt2timestamp(date_time: datetime) -> int:
+    """
+    Converts a python datetime object to unix-timestamp.
+
+    :param date_time: An instance of datetime.
+    :type date_time: datetime
+    :return: Unix timestamp integer.
+    :rtype: int
+    """
+    return int(date_time.timestamp() * 1000)
    @@ -327,5 +342,7 @@

    Source code for dialogy.utils.misc

         
     
     
+    
+    
     
 
\ No newline at end of file
diff --git a/docs/_modules/dialogy/utils/naive_lang_detect.html b/docs/_modules/dialogy/utils/naive_lang_detect.html
index 1745353e..43218f1d 100644
--- a/docs/_modules/dialogy/utils/naive_lang_detect.html
+++ b/docs/_modules/dialogy/utils/naive_lang_detect.html
@@ -5,7 +5,7 @@
     
 
     
-        dialogy.utils.naive_lang_detect - dialogy 0.9.1 documentation
+        dialogy.utils.naive_lang_detect - dialogy 0.9.4 documentation
       
     
     
@@ -124,7 +124,7 @@
    -
    dialogy 0.9.1 documentation
     +
    dialogy 0.9.4 documentation
    @@ -147,7 +147,7 @@
    - dialogy 0.9.1 documentation + dialogy 0.9.4 documentation @@ -160,11 +160,13 @@
  • Input
  • Output
  • Workflow
    • -
  • Base Plugin
    • +
  • Plugin
    • +
  • Intent
    • +
  • Entity
    • +
  • Slots

    • Types

    Plugins

    Misc

      @@ -267,5 +269,7 @@

      Source code for dialogy.utils.naive_lang_detect

      < + + \ No newline at end of file diff --git a/docs/_modules/dialogy/utils/normalize_utterance.html b/docs/_modules/dialogy/utils/normalize_utterance.html index 3fcec026..67aa9486 100644 --- a/docs/_modules/dialogy/utils/normalize_utterance.html +++ b/docs/_modules/dialogy/utils/normalize_utterance.html @@ -5,7 +5,7 @@ - dialogy.utils.normalize_utterance - dialogy 0.9.3 documentation + dialogy.utils.normalize_utterance - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.3 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -161,10 +161,12 @@
    • Output
    • Workflow
    • Plugin
      • +
    • Intent
      • +
    • Entity
      • +
    • Slots

    Types

    Plugins

      @@ -394,7 +395,7 @@

      Source code for dialogy.utils.normalize_utterance

      Convert a list of utterances to a list of alternatives. """ return [ - " ".join([alternative[key] for alternative in alternatives]) + " ".join([alternative[key] for alternative in alternatives]).lower() for alternatives in itertools.product(*utterances) ]
      @@ -405,34 +406,17 @@

      Source code for dialogy.utils.normalize_utterance

      The output will be a list of strings since models will expect that. - .. _normalize: - - .. ipython:: python - - In [1]: from dialogy.utils.normalize_utterance import normalize - In [2]: # A popular case - In [3]: normalize([[{"transcript": "this"}, {"transcript": "works"}]]) .. ipython:: python - In [3]: # A case with multiple utterances - In [4]: normalize([ - [{"transcript": "hello hello?", "transcript": "yellow yellow?"}], - [{"transcript": "I wanted to check"}], - [{"transcript": "if you have space for us?"}] - ]) + from dialogy.utils.normalize_utterance import normalize - .. ipython:: python - - In [5]: normalize([{"transcript": "I wanted to know umm hello?"}]) - - .. ipython:: python - - In [6]: normalize(["I wanted to know umm hello?"]) - - .. ipython:: python + normalize([[{"transcript": "hi"}]]) + normalize([[{"transcript": "hello"}], [{"transcript": "world"}]]) + normalize([{"transcript": "I wanted to know umm hello?"}]) + normalize(["I wanted to know umm hello?"]) + normalize("I wanted to know umm hello?") - In [7]: normalize("I wanted to know umm hello?") :param maybe_utterance: Arbitrary input. :type maybe_utterance: Any @@ -450,13 +434,13 @@

      Source code for dialogy.utils.normalize_utterance

      return utterance2alternatives(maybe_utterance) if is_unsqueezed_utterance(maybe_utterance): - return [alternative[key] for alternative in maybe_utterance] + return [alternative[key].lower() for alternative in maybe_utterance] if is_list(maybe_utterance) and is_list_of_string(maybe_utterance): - return [utterance for utterance in maybe_utterance] + return [utterance.lower() for utterance in maybe_utterance] if is_string(maybe_utterance): - return [maybe_utterance] + return [maybe_utterance.lower()] else: raise TypeError( diff --git a/docs/_modules/dialogy/workflow/workflow.html b/docs/_modules/dialogy/workflow/workflow.html index 02a943f7..725f31fd 100644 --- a/docs/_modules/dialogy/workflow/workflow.html +++ b/docs/_modules/dialogy/workflow/workflow.html @@ -5,7 +5,7 @@ - dialogy.workflow.workflow - dialogy 0.9.3 documentation + dialogy.workflow.workflow - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.3 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -219,131 +219,64 @@

      Source code for dialogy.workflow.workflow

       """
-The :ref:`Workflow<WorkflowClass>` represents a black box for an SLU microservice, and its terminals, :ref:`input <Input>` and :ref:`output <Output>`
-represent request response structure of the SLU API.
+.. _workflow:
 
-.. note::
+A :code:`workflow` is supposed to run tasks that can be anticipated well in advance.
 
-    The :ref:`Input<Input>` has some extra properties that aren't part of the SLU API.
-    We have kept reserves for intermediates, and flattened as most other nested inputs 
-    aren't required and add to bloat.
+Here are few demo's where you can see mock :code:`workflow` in action.
 
-.. attention::
+- :ref:`RuleBasedSlotFillerPlugin<rule_slot_filler>`
+- :ref:`VotePlugin<vote_plugin>`
+- :ref:`DucklingPlugin<duckling_plugin>`
 
-    :ref:`Input<Input>` and :ref:`Output<Output>` instances are immutable.
+A workflow is a hollow conduit, think of a vertically hanging pipe without any medium. If you were to drop a block of ...
+`anything`? through it, it would pass through with a thud on the ground `(yes we assumed gravity)`.
 
-What does it do?
-----------------
+Structure
+----------
 
-Produce the response for a single turn of a conversation. It recieves (not an exhaustive list):
+A workflow allows flexibility, that's why. There is very little structure to it. We have:
 
-- Utterance.
-- Timezone
-- Language code
-- Locale
-- Current State of the conversation
-- Slot tracker
+- input
+- output
+- plugins
 
+Apart from these, we expect at the core, an inference function with machine learning models. Which ones? A N Y ones.
+As long as you have the compute, there is no restriction. Use statistical models or DL or a bunch of conditions,
+your :ref:`Workflow <Workflow>` won't judge you.
 
-as :ref:`inputs<Input>`, parses these, produces a few intermediates and finally returns the :ref:`intent <Intent>` 
-and optionally :ref:`entities <BaseEntity>`. Details will be documented within the specific plugins.
+Advantages
+-----------
 
-.. mermaid::
+The :ref:`Plugin<plugin>` concept takes care of the sauciness 🍅 of this project. Any functionality can be bundled into
+a :ref:`Plugin<plugin>` and they are portable over to foreign workflows. A :ref:`Plugin<plugin>` proxies inputs through
+an :code:`access` function (an argument to every :ref:`Plugin<plugin>`) and relays output through a :code:`mutate`
+function (another argument for every :ref:`Plugin<plugin>`). These two functions define interactions between many many
+(sic) :ref:`Plugins<plugin>` without knowing the inner workings of a :ref:`Workflow <workflow>`.
 
-    classDiagram
-        direction TB
-        Workflow --> "1" Input: has
-        Workflow --> "1" Output: has
-        Workflow --> "many" Plugin: has
+Take a look at :ref:`DucklingPlugin <duckling_plugin>`, this plugin handles inputs, manages the default :code:`json` output
+into neatly bundled :ref:`BaseEntity <base_entity>` and other similar classes. Another plugin
+:ref:`RuleBasedSlotFillerPlugin<rule_slot_filler>` takes care of slot names and the entity types that should be filled
+within.
 
-        class Workflow {
-            Workflow: +Input input
-            Workflow: +Output output
-            Workflow: +set(path: str, value: Any)
-            Workflow: +run(input: Input)
-        }
+If your classifier predicts an :ref:`Intent<intent>` with :ref:`Slots<slot>` supporting any of those entities, then
+slot-filling is not a worry.
 
-        class Input {
-            +List~Utterance~ utterance
-            +int reference_time
-            +str current_state
-            +str lang
-            +str locale
-            +dict slot_tracker
-            +str timezone
-            +dict json()
-        }
-
-        class Output {
-            +List~Intent~ intents
-            +List~BaseEntity~ entities
-            +dict json()
-        }
-
-        <<abstract>> Plugin
-        class Plugin {
-            +str dest
-            *any utility(input: Input, output: Output)
-        }
-
-
-
-How does it do it?
-------------------
-
-- A workflow contains a sequence of :ref:`plugins <AbstractPlugin>`. The sequence is important.
-- The sequence describes the order in which :ref:`plugins <AbstractPlugin>` are run.
-- The plugins can save their output within the workflow's :ref:`input<Input>` or :ref:`output<Output>`.
-- After execution of all plugins, the :ref:`workflow <WorkflowClass>` returns a pair of serialized :ref:`input<Input>` and :ref:`output<Output>`.
-
-
-.. mermaid::
-
-    flowchart TB
-
-        subgraph Workflow
-
-            subgraph input
-                utterance
-                reference_time
-            end
-
-            subgraph output
-                intents
-                entities
-            end
-
-            input --> plugin
-            output --> plugin
-            plugin -->|update| input
-            plugin -->|update| output
-            plugin -->|foreach| plugin
-
-        end
-        output --> output_json
-        input --> input_json
-
-        output_json --> workflow_output
-        input_json --> workflow_output
-
-        subgraph Response
-            output_json
-        end
+The aim of this project is to be largest supplier of plugins for SLU applications.
 
+.. warning:: The :ref:`Workflow<workflow>` class is not supposed to be used as it is. Ideally it should have been an
+    abstract class. There are some design considerations which make that a bad choice. We want methods to be overridden
+    to offer flexibility of use.
 """
-from __future__ import annotations
-
 import copy
 import time
 from threading import Lock
-from typing import Any, Dict, List, Optional, Tuple
+from typing import Any, Dict, List
 
 import attr
 import pandas as pd
 
 from dialogy import constants as const
-from dialogy.base.input import Input
-from dialogy.base.output import Output
 from dialogy.base.plugin import Plugin
 from dialogy.utils.logger import logger
 
@@ -351,9 +284,20 @@ 
      Source code for dialogy.workflow.workflow
       
      [docs]@attr.s
 class Workflow:
     """
-    SLU API blackbox.
+    This is a light but fairly flexible workflow for building a machine learning pipeline.
 
-    .. _WorkflowClass:
+    Requirements
+    - A list of pre-processing functions.
+    - A list of post-processing functions.
+
+    Abstract classes put constraints on method signatures which isn't desired because a couple of methods
+    here could use more arguments, say, `load_model()` requires `path` and `version` and in some other cases
+    `path`, `version` and `language`.
+
+    :param plugins: A list of functions to execute before inference.
+    :type plugins: Optional[List[PluginFn]]
+    :param debug: log level shifts to debug if True.
+    :type debug: bool
     """
 
     plugins = attr.ib(
@@ -361,119 +305,48 @@ 
      Source code for dialogy.workflow.workflow
               type=List[Plugin],
         validator=attr.validators.instance_of(list),
     )
-    """
-    List of :ref:`plugins <AbstractPlugin>`.
-    """
-
-    input: Optional[Input] = attr.ib(
-        default=None,
-        kw_only=True,
-        validator=attr.validators.optional(attr.validators.instance_of(Input)),
-    )
-    """
-    Represents the SLU API request object. A few nested properties and intermediates are flattened out
-    for readability.
-    """
-
-    output: Optional[Output] = attr.ib(
-        default=None,
-        kw_only=True,
-        validator=attr.validators.optional(attr.validators.instance_of(Output)),
-    )
-    """
-    Represents the SLU API response object.
-    """
-
+    input: Dict[str, Any] = attr.ib(factory=dict, kw_only=True)
+    output: Dict[str, Any] = attr.ib(factory=dict, kw_only=True)
     debug = attr.ib(
         type=bool, default=False, validator=attr.validators.instance_of(bool)
     )
-    """
-    Switch on/off the debug logs for the workflow.
-    """
     lock: Lock
-    """
-    A :ref:`workflow <WorkflowClass>` isn't thread-safe by itself. We need locks
-    to prevent race conditions.
-    """
     NON_SERIALIZABLE_FIELDS = [const.PLUGINS, const.DEBUG, const.LOCK]
 
     def __attrs_post_init__(self) -> None:
         """
         Post init hook.
         """
-        self.__reset()
+        self.set_io()
         self.lock = Lock()
 
-    def __reset(self) -> None:
+
      [docs]    def set_io(self) -> None:
         """
         Use this method to keep workflow-io in the same format as expected.
         """
-        self.input = None
-        self.output = Output()
-
-
      [docs]    def set(self, path: str, value: Any, replace: bool = False) -> Workflow:
-        """
-        Set attribute path with value.
+        self.input: Dict[str, Any] = {}
+        self.output: Dict[str, Any] = {const.INTENTS: [], const.ENTITIES: []}
      
 
-        This method (re)-sets the input or output object without losing information
-        from previous instances.
-
-        :param path: A '.' separated attribute path.
-        :type path: str
-        :param value: A value to set.
-        :type value: Any
-        :return: This instance
-        :rtype: Workflow
-        """
-        dest, attribute = path.split(".")
-
-        if dest == "input":
-            self.input = Input.from_dict({attribute: value}, reference=self.input)
-        elif dest == "output" and isinstance(value, list):
-
-            if replace:
-                self.output = Output.from_dict(
-                    {attribute: value}, reference=self.output
-                )
-            else:
-                if attribute == const.INTENTS:
-                    previous_value = self.output.intents  # type: ignore
-                elif attribute == const.ENTITIES:
-                    previous_value = self.output.entities  # type: ignore
-                else:
-                    raise ValueError(
-                        f"Unknown attribute {attribute} tracked on Output."
-                    )
-
-                combined_value = sorted(
-                    previous_value + value,
-                    key=lambda parse: parse.score or 0,
-                    reverse=True,
-                )
-                self.output = Output.from_dict(
-                    {attribute: combined_value}, reference=self.output
-                )
-
-        elif dest == "output":
-            raise ValueError(f"{value=} should be a List[Intent] or List[BaseEntity].")
-        else:
-            raise ValueError(f"{path} is not a valid path.")
-        return self
      
-
-
      [docs]    def execute(self) -> Workflow:
+
      [docs]    def execute(self) -> None:
         """
         Update input, output attributes.
 
         We iterate through pre/post processing functions and update the input and
         output attributes of the class. It is expected that pre-processing functions
         would modify the input, and post-processing functions would modify the output.
+
+        Args:
+            processors (`List`): The list of preprocess or postprocess functions.
+
+        Raises:
+            `TypeError`: If any element in processors list is not a Callable.
         """
         history = {}
         for plugin in self.plugins:
             if not callable(plugin):
                 raise TypeError(f"{plugin=} is not a callable")
 
-            # logs are available only when debug=False during class initialization
+            # logs are available only when debug=True during class initialization
             if self.debug:
                 history = {
                     "plugin": plugin,
@@ -485,7 +358,7 @@ 
      Source code for dialogy.workflow.workflow
                   start = time.perf_counter()
             plugin(self)
             end = time.perf_counter()
-            # logs are available only when debug=False during class initialization
+            # logs are available only when debug=True during class initialization
             if self.debug:
                 history["after"] = {
                     "input": self.input,
@@ -493,10 +366,9 @@ 
      Source code for dialogy.workflow.workflow
                       }
                 history["perf"] = round(end - start, 4)
             if history:
-                logger.debug(history)
-        return self
      
+                logger.debug(history)
      
 
-
      [docs]    def run(self, input_: Input) -> Tuple[Dict[str, Any], Dict[str, Any]]:
+
      [docs]    def run(self, input_: Any) -> Any:
         """
         .. _workflow_run:
 
@@ -504,21 +376,26 @@ 
      Source code for dialogy.workflow.workflow
       
         The current workflow exhibits the following simple procedure:
         pre-processing -> inference -> post-processing.
+
+        Args:
+            input_ (`Any`): This function receives any arbitrary input. Subclasses may enforce
+            a stronger check.
+
+        Returns:
+            (`Any`): This function can return any arbitrary value. Subclasses may enforce a stronger check.
         """
         with self.lock:
             self.input = input_
-            return self.execute().flush()
      
+            self.execute()
+            output = copy.copy(self.output)
+            self.flush()
+        return output
      
 
-
      [docs]    def flush(self) -> Tuple[Dict[str, Any], Dict[str, Any]]:
+
      [docs]    def flush(self) -> None:
         """
         Reset :code:`workflow.input` and :code:`workflow.output`.
         """
-        if self.input is None or self.output is None:
-            return {}, {}
-        input_ = copy.deepcopy(self.input.json())
-        output = copy.deepcopy(self.output.json())
-        self.__reset()
-        return input_, output
      
+        self.set_io()
      
 
 
      [docs]    def json(self) -> Dict[str, Any]:
         """
@@ -532,7 +409,7 @@ 
      Source code for dialogy.workflow.workflow
                   not in self.NON_SERIALIZABLE_FIELDS,
         )
      
 
-
      [docs]    def train(self, training_data: pd.DataFrame) -> Workflow:
+
      [docs]    def train(self, training_data: pd.DataFrame) -> None:
         """
         Train all the plugins in the workflow.
 
@@ -547,11 +424,7 @@ 
      Source code for dialogy.workflow.workflow
                   plugin.train(training_data)
             transformed_data = plugin.transform(training_data)
             if transformed_data is not None:
-                training_data = transformed_data
-        return self
      
-
-
-47
+                training_data = transformed_data
      diff --git a/docs/_modules/index.html b/docs/_modules/index.html index 462a87eb..20d18fd2 100644 --- a/docs/_modules/index.html +++ b/docs/_modules/index.html @@ -5,7 +5,7 @@ - Overview: module code - dialogy 0.9.3 documentation + Overview: module code - dialogy 0.9.4 documentation @@ -124,7 +124,7 @@
      -
      dialogy 0.9.3 documentation
       +
      dialogy 0.9.4 documentation
      @@ -147,7 +147,7 @@
      - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -219,8 +219,6 @@

      All modules for which code is available

      -
      dialogy 0.9.3 documentation
       +
      dialogy 0.9.4 documentation
      @@ -146,7 +146,7 @@
      - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -219,7 +219,7 @@

      Index

      -
      A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | R | S | T | U | V | W | X
      +
      A | B | C | D | E | F | G | I | J | L | M | N | O | P | R | S | T | U | V | W | X

      A

      @@ -237,38 +237,20 @@

      A

    • add_workflow_command_arguments() (in module dialogy.cli)
      • +
    + - @@ -283,12 +265,8 @@

    B

    @@ -303,31 +281,13 @@

    C

  • cleanup() (Intent method)
    • -
  • clear() (Slot method) -
    • -
  • clf_feature (Input attribute) -
    • -
  • collect_datetime_values() (TimeEntity method) - -
    • @@ -337,13 +297,7 @@

    C

    D

    + -
    +
    • dialogy.plugins.text.voting @@ -548,20 +455,11 @@

      D

    • module
    -
    • dialogy.types.entity
      • -
    • - dialogy.types.entity.amount_of_money - -
    • @@ -569,62 +467,6 @@

      D

      • -
    • - dialogy.types.entity.credit_card_number - -
      • -
    • - dialogy.types.entity.deserialize - -
      • -
    • - dialogy.types.entity.duration - -
      • -
    • - dialogy.types.entity.keyword - -
      • -
    • - dialogy.types.entity.numerical - -
      • -
    • - dialogy.types.entity.people - -
      • -
    • - dialogy.types.entity.time - -
      • -
    • - dialogy.types.entity.time_interval - -
    • @@ -674,13 +516,6 @@

      D

      • -
    • - dialogy.utils.datetime - -
    • @@ -734,25 +569,13 @@

      D

  • dict_get() (in module dialogy.utils.normalize_utterance)
    • -
  • dim (BaseEntity attribute) - -
    • -
  • dp_search() (ListSearchPlugin method) -
    • -
  • dt2timestamp() (in module dialogy.utils.datetime) +
  • dt2timestamp() (in module dialogy.utils.misc)
  • DucklingPlugin (class in dialogy.plugins.text.duckling_plugin)
    • -
  • DucklingPluginLB (class in dialogy.plugins.text.lb_plugin) -
    • -
  • DurationEntity (class in dialogy.types.entity.duration) +
  • DURATION (EntityKeys attribute)
    • @@ -762,7 +585,7 @@

    D

    E

    + -
    @@ -855,54 +642,16 @@

    G

    -
    - - -
    -

    H

    - - -
    @@ -912,34 +661,22 @@

    H

    I

    @@ -968,16 +697,10 @@

    I

    J

    -
    -

    K

    - - -
    -
    -

    L

    diff --git a/docs/index.html b/docs/index.html index 97b556b4..f5856502 100644 --- a/docs/index.html +++ b/docs/index.html @@ -6,7 +6,7 @@ - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -242,20 +242,20 @@

    Topics 📄

    1

    -

    Input

    +

    Input

    A fundamental unit that describes the inputs (along with their derivates) of the inference API.

    2

    -

    Output

    +

    Output

    A fundamental unit that describes the output of the inference API.

    3

    -

    Workflow

    +

    Workflow

    A workflow has an Input and an Output. It also contains plugins. The run(...) produces the inference API’s response.

    4

    -

    Plugin

    +

    Plugin

    The base (abstract) class for all plugins.

    [how to write plugins]

    This class requires creation of a utility method of the following signature:

    @@ -266,7 +266,7 @@

    Topics 📄 -

    DucklingPlugin

    +

    DucklingPlugin

    Connects to the Duckling API and returns entities of type:

    1. date

      2. @@ -288,17 +288,17 @@

      Topics 📄input.clf_feature and returns List[Intent].

      -

      CombineDateTimeOverSlots

      +

      CombineDateTimeOverSlots

      This plugin combines date/time entities tracked by the slot filler (previous turns) with time/date entities extracted in the current turn.

      -

      WERCalibrationPlugin

      +

      WERCalibrationPlugin

      This plugin can only be used with the proprietary ASR of skit.ai. Uses lm_score and am_score to detect if the current set of code:input.utterances are too noisy for the intent classifier to use.

      -

      CombineDateTimeOverSlots

      +

      CombineDateTimeOverSlots

      This plugin can only be used with the proprietary ASR of skit.ai. Uses lm_score and am_score to detect if the current set of code:input.utterances are too noisy for the intent classifier to use.

      diff --git a/docs/objects.inv b/docs/objects.inv index 1aac8217..b3da7b8b 100644 Binary files a/docs/objects.inv and b/docs/objects.inv differ diff --git a/docs/py-modindex.html b/docs/py-modindex.html index 9e538389..a1e6fe25 100644 --- a/docs/py-modindex.html +++ b/docs/py-modindex.html @@ -4,7 +4,7 @@ - Python Module Index - dialogy 0.9.3 documentation + Python Module Index - dialogy 0.9.4 documentation @@ -123,7 +123,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -146,7 +146,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -247,18 +247,6 @@

    Python Module Index

    dialogy.base.entity_extractor - - -     - dialogy.base.input - - - - -     - dialogy.base.output - -     @@ -325,54 +313,24 @@

    Python Module Index

    dialogy.plugins.text.classification - - -     - dialogy.plugins.text.classification.mlp - - - - -     - dialogy.plugins.text.classification.tokenizers - -     dialogy.plugins.text.classification.xlmr - - -     - dialogy.plugins.text.combine_date_time - -     dialogy.plugins.text.duckling_plugin - - -     - dialogy.plugins.text.lb_plugin - -     dialogy.plugins.text.list_entity_plugin - - -     - dialogy.plugins.text.list_search_plugin - -     @@ -409,66 +367,12 @@

    Python Module Index

    dialogy.types.entity - - -     - dialogy.types.entity.amount_of_money - -     dialogy.types.entity.base_entity - - -     - dialogy.types.entity.credit_card_number - - - - -     - dialogy.types.entity.deserialize - - - - -     - dialogy.types.entity.duration - - - - -     - dialogy.types.entity.keyword - - - - -     - dialogy.types.entity.numerical - - - - -     - dialogy.types.entity.people - - - - -     - dialogy.types.entity.time - - - - -     - dialogy.types.entity.time_interval - -     @@ -511,12 +415,6 @@

    Python Module Index

    dialogy.utils - - -     - dialogy.utils.datetime - -     diff --git a/docs/search.html b/docs/search.html index 10825b25..b2892cb8 100644 --- a/docs/search.html +++ b/docs/search.html @@ -4,7 +4,7 @@ - Search - dialogy 0.9.3 documentation + Search - dialogy 0.9.4 documentation @@ -122,7 +122,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -145,7 +145,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation diff --git a/docs/searchindex.js b/docs/searchindex.js index 598a97dd..497459a7 100644 --- a/docs/searchindex.js +++ b/docs/searchindex.js @@ -1 +1 @@ -Search.setIndex({docnames:["index","source/dialogy","source/dialogy.base","source/dialogy.base.entity_extractor","source/dialogy.base.input","source/dialogy.base.output","source/dialogy.base.plugin","source/dialogy.cli","source/dialogy.constants","source/dialogy.plugins","source/dialogy.plugins.text","source/dialogy.plugins.text.calibration","source/dialogy.plugins.text.canonicalization","source/dialogy.plugins.text.classification","source/dialogy.plugins.text.combine_date_time","source/dialogy.plugins.text.duckling_plugin","source/dialogy.plugins.text.lb_plugin","source/dialogy.plugins.text.list_entity_plugin","source/dialogy.plugins.text.list_search_plugin","source/dialogy.plugins.text.merge_asr_output","source/dialogy.plugins.text.slot_filler","source/dialogy.plugins.text.voting","source/dialogy.types","source/dialogy.types.entity","source/dialogy.types.entity.amount_of_money","source/dialogy.types.entity.base_entity","source/dialogy.types.entity.credit_card_number","source/dialogy.types.entity.deserialize","source/dialogy.types.entity.duration","source/dialogy.types.entity.keyword","source/dialogy.types.entity.numerical","source/dialogy.types.entity.people","source/dialogy.types.entity.time","source/dialogy.types.entity.time_interval","source/dialogy.types.intent","source/dialogy.types.plugin","source/dialogy.types.signal","source/dialogy.types.slots","source/dialogy.types.utterances","source/dialogy.utils","source/dialogy.workflow","source/modules"],envversion:{"sphinx.domains.c":2,"sphinx.domains.changeset":1,"sphinx.domains.citation":1,"sphinx.domains.cpp":4,"sphinx.domains.index":1,"sphinx.domains.javascript":2,"sphinx.domains.math":2,"sphinx.domains.python":3,"sphinx.domains.rst":2,"sphinx.domains.std":2,"sphinx.ext.viewcode":1,sphinx:56},filenames:["index.rst","source/dialogy.rst","source/dialogy.base.rst","source/dialogy.base.entity_extractor.rst","source/dialogy.base.input.rst","source/dialogy.base.output.rst","source/dialogy.base.plugin.rst","source/dialogy.cli.rst","source/dialogy.constants.rst","source/dialogy.plugins.rst","source/dialogy.plugins.text.rst","source/dialogy.plugins.text.calibration.rst","source/dialogy.plugins.text.canonicalization.rst","source/dialogy.plugins.text.classification.rst","source/dialogy.plugins.text.combine_date_time.rst","source/dialogy.plugins.text.duckling_plugin.rst","source/dialogy.plugins.text.lb_plugin.rst","source/dialogy.plugins.text.list_entity_plugin.rst","source/dialogy.plugins.text.list_search_plugin.rst","source/dialogy.plugins.text.merge_asr_output.rst","source/dialogy.plugins.text.slot_filler.rst","source/dialogy.plugins.text.voting.rst","source/dialogy.types.rst","source/dialogy.types.entity.rst","source/dialogy.types.entity.amount_of_money.rst","source/dialogy.types.entity.base_entity.rst","source/dialogy.types.entity.credit_card_number.rst","source/dialogy.types.entity.deserialize.rst","source/dialogy.types.entity.duration.rst","source/dialogy.types.entity.keyword.rst","source/dialogy.types.entity.numerical.rst","source/dialogy.types.entity.people.rst","source/dialogy.types.entity.time.rst","source/dialogy.types.entity.time_interval.rst","source/dialogy.types.intent.rst","source/dialogy.types.plugin.rst","source/dialogy.types.signal.rst","source/dialogy.types.slots.rst","source/dialogy.types.utterances.rst","source/dialogy.utils.rst","source/dialogy.workflow.rst","source/modules.rst"],objects:{"":[[1,0,0,"-","dialogy"]],"dialogy.base":[[3,0,0,"-","entity_extractor"],[4,0,0,"-","input"],[5,0,0,"-","output"],[6,0,0,"-","plugin"]],"dialogy.base.entity_extractor":[[3,1,1,"","EntityScoringMixin"],[3,4,1,"","entity_scoring"]],"dialogy.base.entity_extractor.EntityScoringMixin":[[3,2,1,"","aggregate_entities"],[3,2,1,"","apply_filters"],[3,2,1,"","entity_consensus"],[3,2,1,"","make_transform_values"],[3,2,1,"","remove_low_scoring_entities"],[3,3,1,"","threshold"]],"dialogy.base.input":[[4,1,1,"","Input"]],"dialogy.base.input.Input":[[4,3,1,"","clf_feature"],[4,3,1,"","current_state"],[4,2,1,"","from_dict"],[4,3,1,"","history"],[4,2,1,"","json"],[4,3,1,"","lang"],[4,3,1,"","latent_entities"],[4,3,1,"","locale"],[4,3,1,"","previous_intent"],[4,3,1,"","reference_time"],[4,3,1,"","slot_tracker"],[4,3,1,"","timezone"],[4,3,1,"","transcripts"],[4,3,1,"","utterances"]],"dialogy.base.output":[[5,1,1,"","Output"]],"dialogy.base.output.Output":[[5,3,1,"","entities"],[5,2,1,"","from_dict"],[5,3,1,"","intents"],[5,2,1,"","json"]],"dialogy.base.plugin":[[6,1,1,"","Plugin"]],"dialogy.base.plugin.Plugin":[[6,2,1,"","prevent"],[6,2,1,"","train"],[6,2,1,"","transform"],[6,2,1,"","utility"]],"dialogy.cli":[[7,4,1,"","add_project_command_arguments"],[7,4,1,"","add_workflow_command_arguments"],[7,4,1,"","main"],[7,0,0,"-","project"],[7,4,1,"","project_command_parser"],[7,0,0,"-","workflow"]],"dialogy.cli.project":[[7,4,1,"","manage_project"],[7,4,1,"","project_cli"]],"dialogy.cli.workflow":[[7,4,1,"","get_workflow"],[7,4,1,"","train_workflow"],[7,4,1,"","workflow_cli"]],"dialogy.constants":[[8,1,1,"","SIGNAL"]],"dialogy.constants.SIGNAL":[[8,3,1,"","NAME"],[8,3,1,"","REPRESENTATION"],[8,3,1,"","STRENGTH"]],"dialogy.plugins":[[10,0,0,"-","text"]],"dialogy.plugins.text":[[11,0,0,"-","calibration"],[12,0,0,"-","canonicalization"],[13,0,0,"-","classification"],[14,0,0,"-","combine_date_time"],[15,0,0,"-","duckling_plugin"],[16,0,0,"-","lb_plugin"],[17,0,0,"-","list_entity_plugin"],[18,0,0,"-","list_search_plugin"],[19,0,0,"-","merge_asr_output"],[20,0,0,"-","slot_filler"],[21,0,0,"-","voting"]],"dialogy.plugins.text.calibration":[[11,0,0,"-","xgb"]],"dialogy.plugins.text.calibration.xgb":[[11,1,1,"","CalibrationModel"],[11,1,1,"","FeatureExtractor"]],"dialogy.plugins.text.calibration.xgb.CalibrationModel":[[11,2,1,"","filter_asr_output"],[11,2,1,"","inference"],[11,2,1,"","predict"],[11,2,1,"","save"],[11,2,1,"","train"],[11,2,1,"","transform"],[11,2,1,"","utility"],[11,2,1,"","validate"]],"dialogy.plugins.text.calibration.xgb.FeatureExtractor":[[11,2,1,"","features"],[11,2,1,"","fit"],[11,2,1,"","transform"]],"dialogy.plugins.text.canonicalization":[[12,1,1,"","CanonicalizationPlugin"],[12,4,1,"","get_entity_type"]],"dialogy.plugins.text.canonicalization.CanonicalizationPlugin":[[12,2,1,"","mask_transcript"],[12,2,1,"","transform"],[12,2,1,"","utility"]],"dialogy.plugins.text.classification":[[13,0,0,"-","mlp"],[13,0,0,"-","tokenizers"],[13,0,0,"-","xlmr"]],"dialogy.plugins.text.classification.mlp":[[13,1,1,"","MLPMultiClass"]],"dialogy.plugins.text.classification.mlp.MLPMultiClass":[[13,2,1,"","get_formatted_gridparams"],[13,2,1,"","get_gridsearch_grid"],[13,2,1,"","inference"],[13,2,1,"","init_model"],[13,2,1,"","load"],[13,2,1,"","save"],[13,2,1,"","train"],[13,2,1,"","utility"],[13,5,1,"","valid_mlpmodel"],[13,2,1,"","validate"]],"dialogy.plugins.text.classification.tokenizers":[[13,4,1,"","identity_tokenizer"]],"dialogy.plugins.text.classification.xlmr":[[13,1,1,"","XLMRMultiClass"]],"dialogy.plugins.text.classification.xlmr.XLMRMultiClass":[[13,2,1,"","inference"],[13,2,1,"","init_model"],[13,2,1,"","load"],[13,2,1,"","save"],[13,2,1,"","train"],[13,2,1,"","utility"],[13,5,1,"","valid_labelencoder"],[13,2,1,"","validate"]],"dialogy.plugins.text.combine_date_time":[[14,1,1,"","CombineDateTimeOverSlots"],[14,4,1,"","has_time_component"],[14,4,1,"","is_date"]],"dialogy.plugins.text.combine_date_time.CombineDateTimeOverSlots":[[14,3,1,"","DATE"],[14,3,1,"","DATETIME"],[14,3,1,"","SUPPORTED_ENTITIES"],[14,3,1,"","TIME"],[14,2,1,"","combine_time_entities_from_slots"],[14,2,1,"","get_tracked_slots"],[14,2,1,"","join"],[14,2,1,"","pick_previously_filled_time_entity"],[14,2,1,"","utility"]],"dialogy.plugins.text.duckling_plugin":[[15,1,1,"","DucklingPlugin"]],"dialogy.plugins.text.duckling_plugin.DucklingPlugin":[[15,3,1,"","FUTURE"],[15,3,1,"","PAST"],[15,2,1,"","apply_entity_classes"],[15,2,1,"","apply_filters"],[15,2,1,"","get_operator"],[15,2,1,"","parse"],[15,2,1,"","select_datetime"],[15,2,1,"","transform"],[15,2,1,"","utility"],[15,2,1,"","validate"]],"dialogy.plugins.text.lb_plugin":[[16,1,1,"","DucklingPluginLB"]],"dialogy.plugins.text.lb_plugin.DucklingPluginLB":[[16,3,1,"","headers"],[16,3,1,"","reference_time"],[16,2,1,"","utility"]],"dialogy.plugins.text.list_entity_plugin":[[17,1,1,"","ListEntityPlugin"]],"dialogy.plugins.text.list_entity_plugin.ListEntityPlugin":[[17,2,1,"","get_entities"],[17,2,1,"","ner_search"],[17,2,1,"","regex_search"],[17,2,1,"","transform"],[17,2,1,"","utility"]],"dialogy.plugins.text.list_search_plugin":[[18,1,1,"","ListSearchPlugin"]],"dialogy.plugins.text.list_search_plugin.ListSearchPlugin":[[18,2,1,"","dp_search"],[18,2,1,"","fuzzy_init"],[18,2,1,"","get_entities"],[18,2,1,"","get_fuzzy_dp_search"],[18,2,1,"","search_regex"],[18,2,1,"","utility"]],"dialogy.plugins.text.merge_asr_output":[[19,1,1,"","MergeASROutputPlugin"],[19,4,1,"","merge_asr_output"]],"dialogy.plugins.text.merge_asr_output.MergeASROutputPlugin":[[19,2,1,"","transform"],[19,2,1,"","utility"]],"dialogy.plugins.text.slot_filler":[[20,0,0,"-","rule_slot_filler"]],"dialogy.plugins.text.slot_filler.rule_slot_filler":[[20,1,1,"","RuleBasedSlotFillerPlugin"]],"dialogy.plugins.text.slot_filler.rule_slot_filler.RuleBasedSlotFillerPlugin":[[20,2,1,"","fill"],[20,2,1,"","utility"]],"dialogy.types":[[23,0,0,"-","entity"],[34,0,0,"-","intent"],[35,0,0,"-","plugin"],[36,0,0,"-","signal"],[37,0,0,"-","slots"],[38,0,0,"-","utterances"]],"dialogy.types.entity":[[24,0,0,"-","amount_of_money"],[25,0,0,"-","base_entity"],[26,0,0,"-","credit_card_number"],[27,0,0,"-","deserialize"],[28,0,0,"-","duration"],[29,0,0,"-","keyword"],[30,0,0,"-","numerical"],[31,0,0,"-","people"],[32,0,0,"-","time"],[33,0,0,"-","time_interval"]],"dialogy.types.entity.amount_of_money":[[24,1,1,"","CurrencyEntity"]],"dialogy.types.entity.amount_of_money.CurrencyEntity":[[24,3,1,"","entity_type"],[24,2,1,"","from_duckling"],[24,2,1,"","get_value"],[24,3,1,"","unit"]],"dialogy.types.entity.base_entity":[[25,1,1,"","BaseEntity"],[25,4,1,"","entity_synthesis"]],"dialogy.types.entity.base_entity.BaseEntity":[[25,2,1,"","add_parser"],[25,3,1,"","alternative_index"],[25,3,1,"","alternative_indices"],[25,3,1,"","body"],[25,2,1,"","copy"],[25,3,1,"","dim"],[25,3,1,"","entity_type"],[25,2,1,"","from_dict"],[25,2,1,"","get_value"],[25,2,1,"","json"],[25,3,1,"","latent"],[25,3,1,"","parsers"],[25,3,1,"","range"],[25,3,1,"","score"],[25,3,1,"","type"],[25,3,1,"","value"],[25,3,1,"","values"]],"dialogy.types.entity.credit_card_number":[[26,1,1,"","CreditCardNumberEntity"]],"dialogy.types.entity.credit_card_number.CreditCardNumberEntity":[[26,3,1,"","entity_type"],[26,2,1,"","from_duckling"],[26,3,1,"","issuer"],[26,3,1,"","value"],[26,3,1,"","values"]],"dialogy.types.entity.deserialize":[[27,4,1,"","deserialize_duckling_entity"]],"dialogy.types.entity.duration":[[28,1,1,"","DurationEntity"]],"dialogy.types.entity.duration.DurationEntity":[[28,2,1,"","as_time"],[28,3,1,"","entity_type"],[28,2,1,"","from_duckling"],[28,3,1,"","normalized"],[28,3,1,"","unit"]],"dialogy.types.entity.keyword":[[29,1,1,"","KeywordEntity"]],"dialogy.types.entity.keyword.KeywordEntity":[[29,3,1,"","alternative_index"],[29,3,1,"","alternative_indices"],[29,3,1,"","body"],[29,3,1,"","dim"],[29,3,1,"","entity_type"],[29,3,1,"","latent"],[29,3,1,"","parsers"],[29,3,1,"","range"],[29,3,1,"","score"],[29,3,1,"","type"],[29,3,1,"","value"],[29,3,1,"","values"]],"dialogy.types.entity.numerical":[[30,1,1,"","NumericalEntity"]],"dialogy.types.entity.numerical.NumericalEntity":[[30,2,1,"","as_time"],[30,3,1,"","entity_type"],[30,2,1,"","from_duckling"]],"dialogy.types.entity.people":[[31,1,1,"","PeopleEntity"]],"dialogy.types.entity.people.PeopleEntity":[[31,3,1,"","entity_type"],[31,2,1,"","from_duckling"]],"dialogy.types.entity.time":[[32,1,1,"","TimeEntity"]],"dialogy.types.entity.time.TimeEntity":[[32,2,1,"","collect_datetime_values"],[32,3,1,"","dim"],[32,2,1,"","from_duckling"],[32,2,1,"","get_value"],[32,3,1,"","grain"],[32,2,1,"","is_uniq_date_from_values"],[32,2,1,"","is_uniq_day_from_values"],[32,3,1,"","origin"],[32,2,1,"","set_entity_type"]],"dialogy.types.entity.time_interval":[[33,1,1,"","TimeIntervalEntity"]],"dialogy.types.entity.time_interval.TimeIntervalEntity":[[33,2,1,"","collect_datetime_values"],[33,3,1,"","dim"],[33,2,1,"","from_duckling"],[33,3,1,"","from_value"],[33,2,1,"","get_value"],[33,3,1,"","origin"],[33,3,1,"","to_value"],[33,3,1,"","type"],[33,3,1,"","value"],[33,3,1,"","value_keys"],[33,3,1,"","values"]],"dialogy.types.intent":[[34,1,1,"","Intent"]],"dialogy.types.intent.Intent":[[34,2,1,"","add_parser"],[34,3,1,"","alternative_index"],[34,2,1,"","apply"],[34,2,1,"","cleanup"],[34,2,1,"","fill_slot"],[34,2,1,"","json"],[34,3,1,"","name"],[34,3,1,"","parsers"],[34,3,1,"","score"],[34,3,1,"","slots"]],"dialogy.types.signal":[[36,0,0,"-","signal"]],"dialogy.types.slots":[[37,1,1,"","Slot"]],"dialogy.types.slots.Slot":[[37,2,1,"","add"],[37,2,1,"","clear"],[37,2,1,"","json"],[37,3,1,"","name"],[37,3,1,"","types"],[37,3,1,"","values"]],"dialogy.utils":[[39,0,0,"-","datetime"],[39,0,0,"-","file_handler"],[39,0,0,"-","logger"],[39,0,0,"-","misc"],[39,0,0,"-","naive_lang_detect"],[39,0,0,"-","normalize_utterance"]],"dialogy.utils.datetime":[[39,4,1,"","dt2timestamp"],[39,4,1,"","is_unix_ts"],[39,4,1,"","make_unix_ts"],[39,4,1,"","unix_ts_to_datetime"]],"dialogy.utils.file_handler":[[39,4,1,"","create_timestamps_path"],[39,4,1,"","load_file"],[39,4,1,"","save_file"]],"dialogy.utils.misc":[[39,4,1,"","traverse_dict"],[39,4,1,"","validate_type"]],"dialogy.utils.naive_lang_detect":[[39,4,1,"","lang_detect_from_text"]],"dialogy.utils.normalize_utterance":[[39,4,1,"","dict_get"],[39,4,1,"","is_each_element"],[39,4,1,"","is_list"],[39,4,1,"","is_list_of_string"],[39,4,1,"","is_string"],[39,4,1,"","is_unsqueezed_utterance"],[39,4,1,"","is_utterance"],[39,4,1,"","normalize"],[39,4,1,"","utterance2alternatives"]],"dialogy.workflow":[[40,0,0,"-","workflow"]],"dialogy.workflow.workflow":[[40,1,1,"","Workflow"]],"dialogy.workflow.workflow.Workflow":[[40,3,1,"","NON_SERIALIZABLE_FIELDS"],[40,3,1,"","debug"],[40,2,1,"","execute"],[40,2,1,"","flush"],[40,3,1,"","input"],[40,2,1,"","json"],[40,3,1,"","lock"],[40,3,1,"","output"],[40,3,1,"","plugins"],[40,2,1,"","run"],[40,2,1,"","set"],[40,2,1,"","train"]],dialogy:[[2,0,0,"-","base"],[7,0,0,"-","cli"],[8,0,0,"-","constants"],[9,0,0,"-","plugins"],[22,0,0,"-","types"],[39,0,0,"-","utils"],[40,0,0,"-","workflow"]]},objnames:{"0":["py","module","Python module"],"1":["py","class","Python class"],"2":["py","method","Python method"],"3":["py","attribute","Python attribute"],"4":["py","function","Python function"],"5":["py","property","Python property"]},objtypes:{"0":"py:module","1":"py:class","2":"py:method","3":"py:attribute","4":"py:function","5":"py:property"},terms:{"0":[4,5,6,8,12,13,14,15,16,17,18,26,28,30,32,33,34,39],"00":[4,14,15,30,32,33],"000":[15,32,33],"0000":14,"01":[15,30,32],"013":[],"014":[],"01t04":[],"01t08":[],"01t09":15,"01t13":15,"02":[15,32,33,39],"02116":13,"02t00":15,"03":[15,33],"034":[],"04":[32,33],"04t18":33,"05":[4,14,15,30,32,33],"05t00":15,"06":[],"07":[],"07t00":33,"07t19":39,"08":[],"08333333333333333":15,"09":[],"099":[],"0th":3,"0x7f01380da490":[],"0x7f0924afb460":[],"0x7f1468ef5460":[],"0x7f1a672044f0":[],"0x7f239778b3d0":[],"0x7f23fdf070d0":[],"0x7f26d4bec4c0":[],"0x7f299944bfa0":[],"0x7f3ba5a21310":[],"0x7f3d1bd58730":[],"0x7f412bc76400":[],"0x7f4509a98af0":[],"0x7f458d2a5580":[],"0x7f46342675b0":[],"0x7f4bd15f0550":[],"0x7f4c9f156460":[],"0x7f500450c1c0":[],"0x7f55c77c84c0":[],"0x7f571b6d4160":[],"0x7f6b03748d00":[],"0x7f6d9c05b2e0":[],"0x7f7147cee370":[],"0x7f7415e69dc0":[],"0x7f7502619eb0":[],"0x7f76760eb550":[],"0x7f7b64f9ffa0":[],"0x7f843a1c5fd0":[],"0x7f8bf9651070":[],"0x7f8ced4426a0":[],"0x7f8fd69c5820":[],"0x7f8fdce11580":[],"0x7f949b167490":[],"0x7f96df591790":[],"0x7fa6eaeb5490":[],"0x7fa9ff656370":[],"0x7fab8d696460":[],"0x7fae7f5a16a0":[],"0x7fbb501c2100":[],"0x7fbc82218610":[],"0x7fbf38c22460":[],"0x7fc28dcf2700":[],"0x7fc8bb68c280":[],"0x7fcf947de4c0":[],"0x7fcf9fc94f40":[],"0x7fd1e3a4fb50":[],"0x7fd38c638df0":[],"0x7fd3a476c3d0":[],"0x7fd6c1eaadf0":[],"0x7fd7ac0ad640":[],"0x7fe08166aa60":[],"0x7fe10623f700":[],"0x7fe13e70e940":15,"0x7fe524f7c6a0":[],"0x7ff4e041f400":[],"0x7ff675de6f40":[],"1":[0,4,5,6,8,13,14,15,17,18,19,26,28,30,31,32,33,34,39],"10":[4,6,14,15,33],"101":[],"105":[],"106":[],"107":[],"108":[],"109":[],"11":[6,15,28,30,39],"110":[],"111":[],"112":[],"113":[],"114":[],"115":15,"116":[],"117":15,"118":[],"119":[],"11t00":[],"11t17":15,"11t18":[],"12":[6,14,15,30],"120":[],"121":[],"122":[],"123":[],"124":[],"125":[],"127":[],"129":[],"12t11":30,"12t15":[],"12t16":[],"12t19":[],"12t20":[],"12t21":[],"13":[6,15,39],"131":[],"134":[],"136":[],"138":[],"13t00":[],"13t06":[],"13t16":[],"13t18":[],"13t22":[],"14":[6,15,39],"14t00":[],"14t16":[],"14t17":[],"14t18":[],"14t19":[],"15":[6,15,39],"15t00":[4,14],"15t06":[],"15t08":[],"15t10":[],"15t11":[],"15t12":[],"15t13":15,"15t15":14,"15t16":[],"15t18":[],"15t19":15,"15t22":[],"15th":14,"16":[6,15,33],"160":[],"163":[],"1644241599537":39,"16t00":[15,32],"16t06":15,"16t16":33,"16t18":33,"16t22":15,"17":[14,15],"17t16":[32,33],"17t18":33,"18":[15,17],"18t16":33,"18t18":33,"19":15,"1911":13,"1996":32,"1_code":4,"1st":[15,32],"2":[0,4,5,6,8,14,15,17,19,26,28,30,31,32,33,34,39],"20":15,"2019":[14,32],"2021":[4,14,15,32,33],"2022":[15,30,32,33,39],"2023":32,"21":[15,32],"210":[],"212":[],"21t00":[15,32,33],"22":[15,30],"23":[15,17],"236":[],"24":[15,17],"25":[15,26],"25t18":33,"26":15,"267":[],"27":15,"270":[],"27t00":32,"27th":32,"28":[15,32],"280":[],"284":[],"28t00":33,"29":15,"29th":32,"2h":[15,28],"2nd":15,"3":[0,4,5,6,14,15,17,19,26,28,30,31,32,33,34,39],"30":[4,14,15,30,32,33,39],"31":15,"31st":15,"31t00":15,"32":15,"320":[],"33":[15,17],"3333333333333333":32,"34":15,"340":[],"344":[],"347":[],"35":15,"350":[],"353":[],"36":15,"367":[],"37":[15,30],"371":[],"372":[],"38":15,"389":[],"39":[15,39],"394":[],"3pm":32,"4":[0,4,6,14,15,17,19,26,28,30,31,32,33,39],"40":15,"404":[],"408":[],"41":[15,26],"411":[],"4111111111111111":[15,26],"42":15,"420":[],"422":[],"43":15,"44":15,"441":[],"444":[],"45":15,"455":[],"459":[],"46":15,"47":15,"470":[],"48":15,"49":[15,39],"499":[],"4pm":32,"5":[4,6,13,15,16,26,28,30,31,32,33,39],"50":39,"504":[],"51":39,"519":[],"52":[3,39],"53":39,"537827":39,"54":[15,39],"542":[],"548":[],"549":[4,6],"55":39,"550":[4,6],"551":[4,6],"552":[4,6],"553":[4,6],"56":[15,39],"57":[],"570":[],"58":[],"59":[],"5pm":[32,33],"5th":15,"6":[4,6,15,30,32,33,39],"609":[],"610":[],"613":[],"625":[],"635":[],"639":15,"64":[],"657":[],"659":[],"66":[],"6666666666666666":[],"667":[],"67":[],"688":[],"6pm":32,"6th":32,"7":[4,6,15,19,30,32,33,39],"705":[],"707":[],"71":[],"72":30,"7200":[15,28],"725":[],"73":[],"733":[],"74":[],"745":[],"750":[],"753":[],"758":[],"769":[],"77":[],"781":[],"784":[],"795":15,"7pm":15,"7th":32,"8":[4,6,14,15,30,32,34,39],"8000":[15,16],"809":[],"81":[],"82":[],"85":[],"86":[],"8601":[4,39],"87":[],"88":[],"884":[],"89":[],"891":[],"893":[],"9":[4,6,15,28,30,39],"90":[],"91":[],"910":[],"914":[],"92":15,"93":[],"94":[],"95":[],"956":[],"958":[],"962":[],"97":[],"98":[],"988":[],"997":[],"9pm":15,"\u0915\u0930":15,"\u0915\u0932":15,"\u0917":15,"\u0925":15,"\u0926":15,"\u0939":15,"abstract":[0,2,11,12,13,17,18,19,20],"break":5,"case":[6,13,14,15,19,28,32,34,39,40],"class":[0,3,4,5,6,8,11,12,13,14,15,16,17,18,19,20,24,25,26,28,29,30,31,32,33,34,37,40],"const":[6,25,39],"default":[4,5,6,7,13,15,24,25,30,39,40],"do":[1,14,15,30,32],"final":40,"float":[3,4,11,15,18,24,25,29,34],"function":[6,7,11,12,13,15,17,18,19,20,25,39,40],"import":[4,5,6,15,17,19,24,26,28,29,30,31,32,33,34,35,38,39,40],"int":[4,13,16,18,25,29,34,39],"long":32,"new":[0,4,5,6,7,17,23,25],"null":14,"return":[0,3,4,5,6,7,11,12,13,14,15,16,17,18,19,20,24,25,26,27,28,30,31,32,33,34,37,39,40],"static":[3,13],"super":6,"switch":[4,40],"true":[4,6,13,14,17,18,20,32,34,39],"try":[4,39],"while":[4,6,13,15,20,25],A:[0,3,4,5,6,7,13,15,16,17,18,19,20,22,25,31,34,39,40],As:20,At:14,For:[32,34],If:[4,5,6,13,15,17,18,20,25,32,34,39],In:[4,5,6,13,15,17,19,26,28,30,31,32,33,34,39],It:[0,4,5,6,11,24,32,40],Its:25,Not:20,Of:15,One:[17,18],Or:39,Such:6,That:5,The:[0,3,4,5,6,7,11,12,13,14,15,17,18,19,20,25,30,34,39,40],Then:4,There:[4,15,18,22,34],These:[3,4,6,15,17,29],To:[6,15,20,32],_:[6,11,13,17,18,19,20,26,28,30,31,32,33,34],__any__:15,__init__:6,__setattr__:[4,6],_alternatives_:11,_callback_:[4,14],_error_:13,_frozen_setattr:[4,6],_greeting_:[5,6],_make:[4,6],_meta:[15,17,28],_oos_:[],_thread:40,ab:13,abc:6,abil:34,abort:7,about:[15,32],abov:[3,15],absenc:4,absolut:15,access:[20,23,24,25,29,35,39],account:[6,34],across:[6,14,15],act:[15,22],action:[32,34],activ:[0,4],activate_latent_ent:15,ad:[3,15,23],adapt:39,add:[25,28,37,40],add_pars:[25,34],add_project_command_argu:7,add_workflow_command_argu:7,addit:[15,20],address:[15,31],adjust_signal_strength:[],advent:15,advis:6,affect:18,after:[6,14,15,28,40],again:6,against:3,aggreg:[3,4],aggregate_ent:3,aggregate_fn:[],ago:[],ahead:28,ai:[0,3,7,15,39],air:22,algorithm:[],all:[0,3,6,10,11,17,18,19,24,25,30,32,33,34,37,40],allocate_lock:40,allow:[4,17,20],along:[0,6],alpha2:15,alreadi:[6,7,15,22],already_paid:15,also:[0,5,6,13,15,17,28,30],altern:[6,11,12,14,15,16,17,18,19,38,39],alternative_index:[4,5,6,14,15,17,24,25,26,27,28,29,30,31,32,33,34],alternative_indic:[15,24,25,26,28,29,30,31,32,33],alwai:[5,15],am:[4,11,15,32,33],am_scor:0,ambigu:15,amount:[0,15,24,34],amount_of_monei:[22,23],an:[0,2,3,5,6,7,11,12,13,15,17,18,19,20,22,25,29,34,39,40],ani:[0,3,4,5,6,7,11,12,13,14,15,17,18,19,20,23,24,25,26,28,29,32,33,34,37,39,40],anoth:[18,34],answer:34,anyon:6,anyth:6,apart:[5,32],api:[0,4,5,15,40],append:[6,13],appli:[3,6,15,20,25,34,39],applic:[0,15],apply_entity_class:15,apply_filt:[3,15],appropri:[13,15],april:32,ar:[0,3,4,5,6,7,11,13,14,15,17,18,19,20,22,23,24,31,32,34,37,39,40],arbitrari:[5,25,39],aren:[6,40],arg:[7,13],argbrighton:[15,16],args_map:13,argument:[0,4,5,7,39],argumentpars:7,arrang:20,artifact:[7,13,40],arxiv:13,as_tim:[28,30],asia:[4,15,26,28,30,31,32,33,39],asr:[0,4,11,13,19,29,39],asr_output:11,assert:39,assign:4,assist:[20,39],associ:[20,22,34],assum:[6,14,15],assumpt:30,attach:[4,6],attempt:[],attent:6,attr:[4,6],attribut:[4,5,7,15,17,24,25,28,30,32,33,34,40],attributeerror:13,automat:34,avail:[7,13,14,17,18,20,34,39],azur:17,bad:11,bank:34,base:[0,1,8,11,12,13,14,15,16,17,18,19,20,24,25,26,28,29,30,31,32,33,34,37,40,41],base_ent:[5,22,23,24,26,28,29,30,31,32,37],baseent:[3,5,6,14,15,16,20,24,25,26,27,28,29,30,31,32,34,37],baseestim:11,bear:[4,5],becaus:[0,3,5,6,15,32],becom:5,been:[6,13,32],beep:39,befor:[0,11,39],being:[25,34],belong:6,below:4,best:19,best_transcript:6,better:34,between:[11,15,20,22,32,33,34],beyond:[],big:15,bit:[11,20],black:40,blackbox:40,bloat:[25,40],block:[0,34],blue:17,bodi:[3,4,14,15,17,24,25,26,28,29,30,31,32,33,34],book:[4,6,15,28],bool:[4,6,11,13,14,17,18,20,25,29,32,34,39],boost:[],booster:[],boot:15,bop:39,bot:14,both:[4,5,32,33],box:40,branch:[0,7],budget:34,build:[0,7],bundl:18,calcul:13,calibr:[9,10],calibra:11,calibrationmodel:11,call:[4,6,15,17,25,34,39],callabl:39,callback_datetim:[4,14],can:[0,3,4,5,6,13,14,15,18,20,22,24,28,30,32,33,34,37,40],candid:[17,18],canonic:[9,10],canonicalizationplugin:12,card:[0,15,23],carri:15,cast:[10,23],categor:34,categori:[6,20],caution:25,cd:[0,7],certain:[15,17,20,39],chang:[0,5,11,25],charact:39,check:[4,6,11,20,25,32,34,39],child:15,children:[15,31],chronolog:32,classif:[9,10,22],classifi:[0,4,11,13,19,20],classmethod:[4,5,24,25,26,28,30,31,32,33],cleanup:[20,34],clear:37,clearli:4,clf_featur:[0,4,6,19],cli:[0,1,41],clone:0,close:6,code:[0,4,15,18,39,40],collect:[6,14,31,32,33],collect_datetime_valu:[32,33],colour:17,column:[6,13],com:[0,3,15,20,39],combin:[0,3,6,14],combine_date_tim:[9,10],combine_time_entities_from_slot:[6,14],combinedatetimeoverslot:[0,4,6,14],combinedatetimeplugin:[],combinetimeentitiesfromslot:[],come:[4,15,19],command:[0,7],command_str:7,common:[4,5],compar:[3,15],con:24,concaten:[0,19,34],conceal:25,conda:0,condit:[6,15,40],condition:15,confid:[3,4,5,6,34],configur:[3,15,18,20,34,39],confus:[4,5],connect:[0,15],consensu:[],consid:[3,11,22],consider:[],consist:[4,15],constant:[1,41],constitu:[],constructor:6,contain:[0,3,4,5,7,11,13,15,20,29,31,32,34,40],content:41,context:15,contextu:[],convers:[4,5,6,22,40],convert:[4,15,25,28,30,34,37,39],copi:[7,25],copier:7,correct:15,correspond:13,could:[4,5,6,15,32],count:[],counter:6,counterpart:[],countri:[4,15],coupl:31,cover:19,cpu:15,creat:[0,4,5,6,7,15,17,20,25,34,39],create_timestamps_path:39,creation:6,credit:[0,15,23],credit_card_numb:[22,23],creditcardnumberent:26,crimson:17,cross:4,csv:7,cue:34,currenc:[15,24],currencyent:24,current:[0,5,6,14,18,20,28,34,37,40],current_ent:14,current_st:[4,6,15,19],d:[4,5,24,26,28,30,31,32,33],dai:[4,14,15,28,30,32],dampen:[],data:[3,4,6,7,11,12,13,15,17,19,25,40],data_column:13,databas:15,datafram:[6,11,12,13,15,17,19,40],dataset:7,date:[0,4,10,14,22,24,28,30,32,33,34,39],date_tim:39,datetim:[1,4,14,15,24,30,32,33,34,41],datetime_filt:[15,16],datetime_slot:34,de:3,debug:[6,11,12,13,14,15,16,17,18,19,20,25,34,40],decemb:[14,15,32],decid:[6,24],decod:15,deep:25,deepli:39,def:[0,4,6],defin:[15,18,20,25,35,39],definit:37,delhi:17,demo:34,depend:[0,3,4,23],depict:[],deploi:15,deriv:[0,4,6,13],descend:[5,6],describ:[0,4,5,6,11,12,13,17,18,19,20,34,39,40],descript:[0,5,19,24,34,40],deseri:[22,23],deserialize_duckling_ent:27,design:20,dest:[6,11,12,13,14,15,16,17,18,19,20,26,28,30,31,32,33],destin:[22,34],destination_path:7,detail:[15,19,34,37,40],detect:[0,29],develop:0,df:11,dialog:[5,14,20],dict:[3,4,5,6,11,13,14,15,16,17,18,19,25,26,28,29,32,33,34,37,39,40],dict_:25,dict_get:39,dict_travers:39,dictionari:[4,5,11,25,32,33,34,37,39],did:32,didn:32,differ:[15,18,25,32,34],digit:39,dim:[24,25,26,28,29,30,31,32,33],dimens:[15,16,24,26,28,30,31,32,33],dir:7,directori:[0,7,39],divis:32,doc:[0,20,34],docker:15,document:40,doe:[1,20],doesn:[3,4,13,25,39],don:[4,5,6,15,34,39],done:[3,5,19],down:15,download:[0,7],dp:18,dp_search:18,drop:15,dry:0,dry_run:39,dsttzinfo:[],dt2timestamp:[15,30,39],duckl:[0,4,14,15,16,18,23,24,30,32],duckling_entity_dict:27,duckling_plugin:[9,10,16,26,28,30,31,32,33],ducklingplugin:[0,5,14,15,16,18,26,28,30,31,32,33],ducklingpluginlb:16,due:39,duplic:3,durat:[0,15,22,23,30,32],duration_cast_oper:[15,27,28],durationent:28,dure:11,e:4,each:[4,18,19,20,25,32,34,39],eagerli:20,easi:17,easili:[],either:[15,22,32,33],elect:[],element:[5,25,39],els:[6,13,32,39],emerald:17,emit:[],en:[4,6,19],en_core_web_sm:17,en_gb:15,en_in:[4,6,15,16,19,26,28,30,31,32,33],en_u:15,enabl:13,encod:39,end:[4,14,15,17,19,26,28,30,32,33],end_dat:34,english:[4,20],enough:29,ensur:0,entir:[4,6],entit:29,entiti:[0,1,3,4,5,6,12,14,15,16,17,18,20,22,34,37,39,40],entity_column:12,entity_consensu:3,entity_dict:15,entity_extractor:[1,2,15,17,18],entity_pattern:18,entity_scor:3,entity_synthesi:25,entity_typ:[4,14,15,17,18,24,25,26,28,29,30,31,32,33],entity_type_val_group:3,entity_type_value_group:3,entityscoringmixin:[3,15,17,18],entri:7,env:[4,6],equal:15,equival:6,error:39,especi:15,etc:[15,29,32],evalu:6,even:[3,4,5,6,20],everi:[3,4],everyth:39,evid:[],ex:32,exactli:4,exampl:[6,7,10,11,20,22,31,32,34],except:39,exclus:39,execut:[6,15,20,34,40],exhaust:40,exhibit:[7,40],exist:[0,2,5,6,7,13,39],exit:[0,7],expect:[4,6,7,13,15,16,18,19,20,22,25,28,39,40],experi:15,explicit:32,explicitli:23,explor:[6,34],expos:7,express:[0,6,15],extra:[15,40],extract:[0,17,18,20,29,37],f:39,facilit:39,fail:39,fairli:0,fallback:4,fallback_int:[],fallback_label:13,fals:[4,6,7,11,12,13,14,15,16,17,18,19,20,24,25,26,28,29,30,31,32,33,34,39,40],featur:[0,4,6,11,13,19,34],featureextractor:11,femal:15,few:[0,6,17,18,20,22,40],field:6,file:[4,6,7,39],file_handl:[1,41],file_nam:39,file_path:39,fill:[4,5,6,10,14,22,34,37],fill_multipl:[20,34],fill_slot:34,filler:[0,20],filter:[3,4,10,11,25],filter_asr_output:11,filter_typ:15,find:[],fine:39,first:[6,11,14],fit:11,fix:32,flag:[13,17,18],flatten:[39,40],flight:[4,15,32,33],flush:40,fn:7,fname:11,follow:[0,4,5,6,7,34,40],forc:0,forest:[],forgot:6,fork:15,form:39,format:[4,11,13,15,32,33,39],former:6,found:[14,20,34],free:15,frequenc:[],from:[3,4,5,6,7,11,13,14,15,17,18,19,24,25,26,28,29,30,31,32,33,34,39,40],from_dict:[4,5,6,18,25],from_duckl:[24,26,28,30,31,32,33],from_valu:[15,33],froze:15,frozen:[2,6],frozeninstanceerror:[4,6],fruit:30,full:6,fun:15,fundament:0,futur:15,fuzzi:18,fuzzy_dp_config:18,fuzzy_init:18,fuzzy_threshold:18,gain:17,gap:[],gener:[6,7,22,32],geographi:15,get:[6,13,15,25,28,32,39,40],get_ent:[17,18],get_entity_typ:12,get_formatted_gridparam:13,get_fuzzy_dp_search:18,get_gridsearch_grid:13,get_oper:15,get_tracked_slot:14,get_valu:[24,25,32,33],get_workflow:7,get_workflow_fn:7,getter:24,girl:15,git:[0,7],github:[0,3,7,15,39],gitlab:[0,7],give:14,given:[6,15,25,39],goe:[],gpe:17,grain:[4,14,15,30,32,33],graph:4,greater:15,green:17,greet:6,grid:13,gridsearch:13,gridsearchcv:13,group:3,guard:[2,10,11,12,13,14,16,17,18,19,20],h:[0,7],ha:[0,3,4,6,13,15,17,25,32,33,34,40],hack:18,had:6,handi:4,handl:[20,24,29,30],happen:15,hard:0,has_time_compon:14,have:[3,4,6,11,14,15,17,20,24,30,31,32,33,34,39,40],head:[0,7],header:16,healthi:[],hello:[4,6,7,39],help:[0,4,6,7,17,25,28,34],here:[6,7,15,32,39],heurist:2,heuristic_based_int:6,heuristicbasedintentplugin:6,hi:4,hi_in:15,high:[],higher:[0,3,34],highest:[3,5],highli:6,hindi:15,histori:[4,6,19],hold:34,hope:39,hour:[15,28,30,32,33],how:[0,1,15],howev:6,html:20,http:[3,4,13,15,16,20,39],hyperparamet:13,hypothesi:4,i:[6,11,15,17,28,31,32,33,34,39],id:15,ideal:[],ident:[3,39],identifi:[4,20,25,34],identity_token:13,immut:[4,5,6,40],impact:18,imper:[],implement:[6,12,15],impli:[],improv:[],in_:6,inaud:11,includ:25,index:[3,34,39],indexerror:[25,39],indic:3,indigo:17,individu:18,infer:[0,11,13,32,40],info:[],inform:[15,25,28,36,40],init_model:13,initi:[13,18],input:[0,1,2,6,11,12,13,14,15,17,18,19,20,26,28,30,31,32,33,39,40],input_:[0,6,15,16,18,19,39,40],input_column:[6,11,12,14,15,16,17,18,19],input_i:4,input_s:3,input_x:4,insert:37,instal:7,instanc:[2,5,6,7,15,25,34,37,39,40],instead:[0,4,6,15,34,39],integ:39,integr:23,intend:25,intent:[0,1,2,4,5,10,11,13,17,19,20,22,26,28,30,31,32,33,37,40],intent_detection_slot_fil:20,intent_vot:[],interact:14,intermedi:[6,18,40],interpret:34,intersect:15,interv:[15,23,24,30],invers:15,invoc:[0,7],involv:18,ip:6,ipython:15,irrespect:20,is_dat:14,is_each_el:39,is_list:39,is_list_of_str:39,is_str:39,is_uniq_date_from_valu:32,is_uniq_day_from_valu:32,is_unix_t:39,is_unsqueezed_utter:39,is_upd:7,is_utter:39,isn:[4,13,15,40],iso3166:15,iso:[4,15,32,33,39],issu:[2,3,15,23,39],issuer:[15,26],ist:[],item:[0,6,20],iter:40,its:[3,6,7,15,20,34,40],itself:[15,40],jade:17,jan:[15,32],job:11,join:[14,19],json:[3,4,5,14,15,25,34,37,40],just:[5,11,15,18,39],k:39,keep:[4,6,18],kei:[4,5,11,13,19,33,37,39],kept:40,key_error:39,keyerror:[25,39],keyword:[2,17,22,23],keywordent:[17,18,20,29],kid:31,knew:6,know:[15,39],known:[23,34],kolkata:[4,15,26,28,30,31,32,33,39],kwarg:[6,7,13],label:[13,17,18],label_column:13,label_count:13,labelencod:13,lack:0,lambda:[6,15,39],lang:[4,6,18,19],lang_detect_from_text:39,languag:[4,15,16,18,26,28,30,31,32,33,39,40],last:[4,6,32,39],latenc:15,latent:[4,15,24,25,26,28,29,30,31,32,33],latent_ent:[4,6,19],latest:[0,7],latter:6,lb_plugin:[9,10],lead:39,learn:[6,11],least:[4,6,13],leav:39,length:3,less:3,lesser:15,let:[6,15,20],li:[0,7],lib:[4,6,39],librari:0,like:[4,5,6,15,22,28,34],line:7,link:[7,20],lint:7,list:[0,3,4,5,6,11,12,13,14,15,16,17,18,19,20,25,26,29,32,33,34,37,39,40],list_entity_plugin:[9,10],list_of_ent:15,list_of_iso_639:4,list_of_tz_database_time_zon:4,list_search_plugin:[9,10],listentityplugin:[5,17],listsearchplugin:[5,18],lm:11,lm_score:0,load:[13,17,39,40],load_fil:39,loader:39,local:[4,6,7,15,16,19,26,28,30,31,32,33,39,40],locat:[7,34],lock:40,log:[4,5,6,34,40],logger:[1,41],logic:39,look:[6,15,34,39],lookup:29,lose:40,lost:[4,34],lower:[0,3],machin:[15,34],made:[],mai:[3,4,6,15,20,32,39],main:7,make:[0,3,7,13,15,28,39],make_transform_valu:3,make_unix_t:[4,39],male:15,man:15,manag:[0,5,40],manage_project:7,mani:[],manner:32,map:[17,18,20,34],mar:39,march:15,mark:19,mask:12,mask_token:12,mask_transcript:12,master:[0,7,15],mastercard:39,match:[15,17,18,20,39],match_dict:18,matchtyp:[17,18],matter:32,mayb:34,maybe_utter:39,me:15,mean:[4,5,6,15,22,34],meant:[25,34],mention:32,merg:19,merge_asr_output:[9,10],merge_asr_output_plugin:[6,19],mergeasroutput:0,mergeasroutputplugin:[6,19],messag:[0,7],met:[],meta:[28,29],metadata:3,method:[0,6,11,12,13,15,16,17,18,19,20,25,30,40],microservic:40,might:[4,6],migrat:0,millisecond:[15,16],miniconda3:[4,6],minimum:3,minut:32,misc:[1,41],mismatch:39,miss:[4,7,19,39],mixin:3,mlp:[9,10],mlpclassifi:13,mlpmulticlass:[5,13,34],mode:39,model:[0,11,13,17,18,39],model_dir:13,model_nam:11,modif:[6,25,34],modifi:[3,25,34,40],modul:[0,41],moment:14,mondai:[15,32],monei:[0,15,24,34],month:[30,32],more:[0,13,15,20,25,34],morn:32,most:[3,4,6,39,40],mother:10,ms:[4,15,39],multipl:[3,15,34,39],must:[5,7,34,39],mutat:[20,25],my:[11,15,26,34],n:[4,13,19,31,39],naiv:6,naive_lang_detect:[1,41],name:[0,4,5,6,7,8,14,15,20,25,34,37,39],namespac:[0,7],natur:[15,16,26,28,30,31,32,33],navi:17,navig:39,need:[0,4,5,6,14,15,17,18,20,25,28,30,32,33,39,40],neg:32,ner:[10,18],ner_search:17,nest:[39,40],never:6,newlin:39,night:32,nlp:[17,18],nlpprogress:20,node:4,noisi:0,nomenclatur:[4,5],non:39,non_serializable_field:40,none:[3,4,5,6,7,11,12,13,14,15,16,17,18,19,20,24,25,26,27,28,29,30,31,32,33,34,39,40],nonetyp:7,normal:[4,15,19,28,39],normalize_utter:[1,41],note:[4,15],noth:[4,5,24,25,26,28,29,30,31,32,33,34,37,40],notic:[15,30],notion:15,now:[4,5,15],np:[],ns:[],number:[0,3,6,13,15,20,22,23,24,28,30,31,32,34],numer:[15,22,23,24,31],numericalent:30,o:[6,15,39],obj:39,obj_typ:39,object:[3,4,5,6,7,8,13,15,25,34,37,39,40],observ:[],obtain:[15,24,26,28,30,31,32,33,34],occurr:[],oct:32,octob:32,off:[4,40],offer:6,often:[15,17],onc:20,one:[3,6,14,20,22,32,40],onli:[0,5,15,17,18,19,20,25,32,34],oop:6,op:40,open:[23,39],oper:[19,28,34],oppos:18,option:[0,3,4,5,6,7,13,14,15,16,17,18,20,24,25,26,29,30,31,32,33,34,39,40],orchestr:20,order:[0,5,6,32,40],org:[4,13],organ:[0,7],origin:[15,22,25,29,30,32,33,34],other:[0,3,4,6,15,17,18,20,25,40],otherwis:13,our:[4,6,15,32,34],out:[4,5,6,15,17,19,26,28,30,31,32,33,34,39,40],output:[0,1,2,6,11,12,13,14,15,16,17,18,19,20,24,26,28,30,31,32,33,39,40],output_column:[6,11,12,14,15,16,17,18,19],output_dir:7,over:[3,6,17],overal:15,overrid:[25,40],overridden:[],packag:41,pai:[6,15],paid:15,pair:[32,40],panda:13,param:[13,18,39],paramet:[3,4,5,6,7,11,12,13,14,15,16,17,18,19,20,24,25,32,33,34,39,40],pari:4,park:32,pars:[4,14,15,16,17,18,24,26,28,30,31,32,33,34,39,40],parser:[4,5,6,7,14,15,17,24,25,26,28,29,30,31,32,33,34],part:[3,13,15,20,32,40],partial:39,particip:22,pass:[0,15,17,20,32,33],passeng:[22,34],past:15,path:[6,7,39,40],pattern:[4,5,15,17,18],pay_lat:15,payment:15,pd:[11,13,15,17,40],peopl:[0,15,22,23,34],peopleent:[15,31],per:[3,4,15,20],percentag:[],person:15,pertain:20,pick:[3,25],pick_previously_filled_time_ent:14,pinpoint:32,pip:0,pipelin:[11,13],piplin:11,pizza:15,pkl:11,place:[5,17],placehold:[4,20,34],plain:[],plan:2,planet:39,plugin:[0,1,2,22,23,25,34,40,41],pluginfn:[20,35],pm:[4,14,15,19,33],poetri:[0,7],point:[3,4,7,15],poor:[],popular:39,posit:[0,7,19,32,34],possibl:[32,34,39],post:40,postprocessor:25,pprint:15,pr:0,practic:20,pre:[20,35,40],predict:[4,5,6,11,13,15,19,34],prefer:[25,34,39],prepar:6,presenc:[3,6,29],present:[4,6,20],preset:20,pretend:7,prevent:[6,15,18,25,40],previou:[0,4,6,14,32,40],previous:[4,14],previous_ent:14,previous_int:[4,6,19],print:6,pristin:18,privat:15,pro:24,problem:[3,6,10,34],procedur:[34,40],process:[5,40],produc:[0,4,5,6,15,16,40],product:[4,25,34],program:39,progress:[22,25,34],project:[1,8,17,22,41],project_cli:7,project_command_pars:7,promin:19,prop:39,proper:13,properti:[4,13,25,34,39,40],property_:25,proport:[],proprietari:0,provid:[6,7,8,11,13,14,15,19,20,23,24,25,26,28,29,30,31,32,33,35,36,38,39],purpos:[5,7,13,25],py:[4,6,39],python3:[4,6],python:[39,40],pythoncod:39,pytz:[15,30],quarter:32,queri:18,r:39,race:40,rais:[4,6,13,19,25,39],rang:[4,14,15,17,24,25,26,28,29,30,31,32,33,39],rank:[0,6],raspberri:17,re:[4,40],reach:15,read:[0,6,13,15,37],readabl:40,readi:[],real:34,rebuild:6,receiv:[14,15],recent:[4,6,39],reciev:40,recurs:39,red:17,reduc:[3,4,39],ref:[],refactor:18,refer:[4,5,6,15,16,24,25,39],reference_dt:[15,30],reference_tim:[4,6,15,16,19,27,30,39],reference_time_column:[15,16],reference_unix_t:[28,30],reftim:[15,16],regex:[10,18,29],regex_search:17,regexflag:[17,18],region:4,rel:[4,15],relat:20,relationship:[20,22],relev:[5,34],remain:[],remov:[3,11,15,20,34,37],remove_low_scoring_ent:3,repeat:15,replac:[6,25,30,40],replace_output:[6,14,20],repositori:0,repres:[4,5,19,30,31,32,33,34,40],represent:[8,25],request:40,requir:[0,4,5,6,15,17,18,22,40],reserv:[22,40],reset:40,resid:39,resolut:15,resolv:15,resort:[],respect:[19,22],respons:[0,15,39,40],rest:6,result:[6,40],retain:[0,24],reus:2,roberta:13,role:20,root:[0,7],round:22,row:11,rtype:[18,25,39],rubi:17,rule:[20,34],rule_slot_fil:[9,10],rulebasedslotfillerplugin:[6,20,34],run:[0,6,15,16,17,19,20,26,28,30,31,32,33,40],s:[0,3,4,5,6,7,11,12,13,15,16,17,18,19,20,28,34,39,40],safe:[25,39,40],safest:[],sai:[5,6,15,32],said:[28,34],same:[3,11,18,34],saniti:11,saturdai:32,save:[6,11,13,15,39,40],save_fil:39,scaffold:7,schema:20,score:[3,4,5,6,11,14,15,17,24,25,26,28,29,30,31,32,33,34],score_round_off:13,search:[3,10,15,18],search_regex:18,searchstrategyclass:18,second:[4,15,28,32],section:[0,6,15,22,25],see:[6,14,15],seem:[6,24],seen:17,select:[3,15],select_datetim:15,self:[0,4,5,6,32],sentenc:[19,22,28,31,32],separ:[4,5,6,19,34,40],sequenc:40,serial:[2,12,40],server:0,servic:[15,18],session:15,set:[0,2,3,4,6,11,14,15,17,18,20,22,34,40],set_entity_typ:32,setup:[0,15],seven:32,seventi:30,shall:20,shape:18,share:[3,34],sharp:11,she:28,ship:[6,18,39],shirt:17,should:[0,4,6,11,13,15,25,34,40],show:[0,6,7,17,18],signal:[1,8,22],signatur:[0,20],signific:[0,17,20,22],similar:[],simpl:[0,7,39,40],simul:15,sinc:[4,5,6,14,32,34,39],singl:[4,5,14,19,20,34,40],sit:11,site:[4,6],size:[],skip:[13,25],skip_entity_attr:25,skip_label:13,skit:0,sklearn:[11,13],slot:[0,1,4,5,6,10,14,15,22,34,40],slot_fil:[9,10],slot_track:[4,6,14,19],slot_typ:34,slow:15,slu:[0,4,5,11,17,22,34,40],small_talk:15,smallest:32,smoothli:0,snapshot:6,so:[6,14,18,22],solv:34,some:[4,6,25,39,40],someon:15,soon:5,sort:[5,15],sourc:[3,4,5,6,7,8,11,12,13,14,15,16,17,18,19,20,24,25,26,27,28,29,30,31,32,33,34,37,39,40],space:[18,39],spaci:[10,18],spacy_nlp:[17,18],sparingli:39,speaker:39,special:34,specif:[25,29,39,40],specifi:[7,15],split:6,spuriou:13,stai:17,stall:15,standard:[15,39],start:[4,7,14,15,17,19,26,28,30,32,33],start_dat:34,state:[4,6,13,15,40],state_column:13,state_expecting_numb:6,std:[],store:[32,33],str:[3,4,5,6,7,11,12,13,14,15,16,17,18,19,24,25,26,28,29,30,31,32,33,34,37,39,40],strategi:[18,19],strength:8,string:[3,4,13,15,17,18,19,32,33,39],strong:[],structur:[0,3,4,18,19,25,40],style:[17,18],subclass:[7,25],submodul:[1,9,10,22,41],subpackag:41,subsequ:14,subset:[4,5],sugar:[25,39],suitabl:40,sundai:32,support:[0,7,15,19,23,34],supported_ent:14,suppos:[],sure:[13,15],sy:15,syntat:39,syntax:4,system:[4,20],t:[3,4,5,6,13,15,25,32,34,39,40],tag:[0,7,11],take:[6,7],task:[17,19],teal:17,tell:[15,20,22,28,32,34],templat:[0,7],tempor:10,temporal_int:15,tend:15,termin:[19,40],test:10,text:[1,9,34,39],tfidf:[11,13],than:[3,15,20],thei:[14,32,34],them:[6,20,40],therefor:32,thi:[0,3,4,5,6,7,11,12,13,14,15,17,18,19,20,22,24,25,26,28,29,30,31,32,33,34,36,37,39,40],third:20,those:[15,34],thread:40,threshold:[3,6,11,12,13,15,17,18],through:[20,40],throughtout:8,ticket:22,ticket_book:34,till:32,time:[0,4,6,10,14,16,18,22,23,24,28,30,34],time_interv:[22,23],timeent:[14,15,20,28,30,32,33],timeinterv:15,timeintervalent:33,timeout:[15,16],timestamp:[4,28,39],timezon:[4,6,15,16,19,26,27,28,30,31,32,33,39,40],to_time_ent:[],to_valu:[15,33],todai:[15,32,33],togeth:32,token:[9,10,18,20,22,29],tommorrow:15,tomorrow:[4,14,15,32],tonight:32,too:0,tool:18,top:[],total:15,traceback:[4,6,39],track:[0,3,4,14,15],tracked_slot:14,tracker:[14,40],train:[0,6,7,11,13,15,17,34,40],train_workflow:7,trainabl:[13,40],training_data:[6,11,12,13,15,17,19,40],transact:34,transcript2int:6,transcript2intentplugin:6,transcript:[3,4,6,11,12,13,15,17,18,19,34,39],transcrit:11,transform:[6,7,11,12,15,17,19,39],transformermixin:11,trascript:11,travel:22,travers:39,traverse_dict:39,tree:[],trial:[],trigger:5,trigger_int:14,trip:22,triplet:31,ts:39,tupl:[3,11,15,16,18,39,40],turn:[0,4,5,14,15,40],twenti:30,two:[11,15,20,28,30,31],type:[0,1,3,4,5,6,7,11,12,13,14,15,16,17,18,19,20,39,40,41],type_:39,typeerror:[19,39],tz:[15,39],tzinfo:[15,30],umm:39,understand:[0,15,34],union:[4,39],uniqu:[3,32],unit:[0,15,24,28,31,32],unix:[4,39],unix_ts_to_datetim:39,unless:[4,20],unsaf:25,unset:39,up:15,updat:[0,2,7,25,34,39,40],upon:22,url:[15,16],us:[0,4,6,7,8,13,14,15,17,18,19,20,22,24,25,28,29,30,32,34,39],usag:[0,7],use_cuda:13,use_lat:15,use_mast:7,use_st:13,use_transform:[6,11,12,14,15,16,17,18,19],user:[0,7,14,15,32],usual:15,utc:[4,6,15,19,27,39],utf:39,util:[0,1,4,6,7,11,12,13,14,15,16,17,18,19,20,25,30,34,41],utter:[0,1,4,6,11,14,15,17,19,22,26,28,30,31,32,33,34,39,40],utterance2altern:39,valid:[5,11,13,15,40],valid_labelencod:13,valid_mlpmodel:13,validate_typ:39,valu:[3,4,6,11,12,13,14,15,17,18,19,20,24,25,26,28,29,30,31,32,33,34,37,39,40],value_kei:33,valueerror:13,variabl:13,variant:31,variou:[15,39],vc:7,vcs_ref:7,vector:11,verif:15,vernacular:[3,7,15,39],version:0,via:[4,15,17,20,26,28,30,31,32,33],virtualenv:0,visa:[15,26],visit:14,vote:[9,10,34],vote_sign:[],voteintentplugin:[],voteplugin:[],w:39,wa:[4,6,15,28,32,34,39],wai:[4,15,22],wake_up_alarm:34,walkthrough:23,wall:15,wallet:34,want:[4,11,14,15,17,28,34,39],wasn:6,we:[0,3,4,5,6,7,11,14,15,17,18,19,20,22,23,24,28,30,32,34,39,40],weak:[],week:32,weekdai:32,weekend:33,weight:[],well:[5,15,32],went:32,wercalibrationplugin:0,were:[4,15,20,25,34],what:[1,14,15,34],when:[14,15,34,39],where:[0,3,6,7,15,28,29,31,32,34],which:[3,5,15,17,18,19,24,25,28,30,31,32,33,34,39,40],whole:31,wiki:4,wikipedia:4,win:[],wish:28,within:[3,4,5,6,15,17,18,22,25,34,37,39,40],without:40,word:6,word_frequ:6,work:[0,4,5,19,22,39],workflow:[0,1,6,11,12,13,15,16,17,18,19,20,23,41],workflow_cli:7,world:[4,7],worst:[],would:[4,5,13,14,15,19,22,34,40],wrap:39,wrapper:17,write:[0,2,4,39],writer:39,wrote:6,x:39,xgb:[9,10],xgboostregressor:11,xlm:13,xlmr:[9,10],xlmrmulticlass:[0,5,13,34],y:11,year:32,yeild:34,yellow:[6,39],yesterdai:[15,32],yet:24,yield:[],you:[0,4,6,14,15,22,39],your:[0,6]},titles:["Dialogy","dialogy package","dialogy.base package","dialogy.base.entity_extractor package","dialogy.base.input package","dialogy.base.output package","dialogy.base.plugin package","dialogy.cli package","dialogy.constants package","dialogy.plugins package","dialogy.plugins.text package","dialogy.plugins.text.calibration package","dialogy.plugins.text.canonicalization package","dialogy.plugins.text.classification package","dialogy.plugins.text.combine_date_time package","dialogy.plugins.text.duckling_plugin package","dialogy.plugins.text.lb_plugin package","dialogy.plugins.text.list_entity_plugin package","dialogy.plugins.text.list_search_plugin package","dialogy.plugins.text.merge_asr_output package","dialogy.plugins.text.slot_filler package","dialogy.plugins.text.voting package","dialogy.types package","dialogy.types.entity package","dialogy.types.entity.amount_of_money package","dialogy.types.entity.base_entity package","dialogy.types.entity.credit_card_number package","dialogy.types.entity.deserialize package","dialogy.types.entity.duration package","dialogy.types.entity.keyword package","dialogy.types.entity.numerical package","dialogy.types.entity.people package","dialogy.types.entity.time package","dialogy.types.entity.time_interval package","dialogy.types.intent package","dialogy.types.plugin package","dialogy.types.signal package","dialogy.types.slots package","dialogy.types.utterances package","dialogy.utils package","dialogy.workflow package","dialogy"],titleterms:{"abstract":6,"do":[4,5,40],all:[4,5,15],amount_of_monei:24,an:4,base:[2,3,4,5,6],base_ent:25,calibr:11,canonic:12,card:26,cast:[15,30],classif:[13,34],cli:7,combine_date_tim:14,constant:8,content:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],contributor:0,creation:0,credit:26,credit_card_numb:26,date:15,datetim:39,deseri:27,dialogi:[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41],doe:40,duckling_plugin:15,durat:28,entiti:[23,24,25,26,27,28,29,30,31,32,33],entity_extractor:3,exampl:15,exist:4,file_handl:39,fill:20,filter:15,frozen:4,get:0,guard:[6,15],heurist:6,how:40,i:[4,5],index:0,input:[4,5],instal:0,instanc:4,integr:[26,28,30,31,32,33],intent:[6,15,34],intent_vot:[],interv:33,issu:4,keyword:[6,29],lb_plugin:16,list_entity_plugin:17,list_search_plugin:18,logger:39,merge_asr_output:19,misc:39,mlp:13,modul:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],mother:15,naive_lang_detect:39,ner:17,normalize_utter:39,number:26,numer:30,output:[4,5],packag:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],peopl:31,plan:6,plugin:[4,5,6,9,10,11,12,13,14,15,16,17,18,19,20,21,26,28,30,31,32,33,35],problem:15,project:[0,7],regex:17,reus:4,rule_slot_fil:20,search:17,see:[4,5],serial:[4,5],set:5,signal:36,slot:[20,37],slot_fil:20,spaci:17,start:0,submodul:[7,11,13,20,36,39,40],subpackag:[1,2,9,10,22,23],tempor:15,test:[0,15],text:[10,11,12,13,14,15,16,17,18,19,20,21],time:[15,32,33],time_interv:33,token:13,topic:0,type:[22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],updat:[4,6],util:39,utter:38,vote:21,walkthrough:[26,28,30,31,32,33],what:40,why:[4,5],workflow:[7,26,28,30,31,32,33,40],write:6,xgb:11,xlmr:13}}) \ No newline at end of file +Search.setIndex({docnames:["index","source/dialogy","source/dialogy.base","source/dialogy.base.entity_extractor","source/dialogy.base.input","source/dialogy.base.output","source/dialogy.base.plugin","source/dialogy.cli","source/dialogy.constants","source/dialogy.plugins","source/dialogy.plugins.text","source/dialogy.plugins.text.calibration","source/dialogy.plugins.text.canonicalization","source/dialogy.plugins.text.classification","source/dialogy.plugins.text.combine_date_time","source/dialogy.plugins.text.duckling_plugin","source/dialogy.plugins.text.lb_plugin","source/dialogy.plugins.text.list_entity_plugin","source/dialogy.plugins.text.list_search_plugin","source/dialogy.plugins.text.merge_asr_output","source/dialogy.plugins.text.slot_filler","source/dialogy.plugins.text.voting","source/dialogy.types","source/dialogy.types.entity","source/dialogy.types.entity.amount_of_money","source/dialogy.types.entity.base_entity","source/dialogy.types.entity.credit_card_number","source/dialogy.types.entity.deserialize","source/dialogy.types.entity.duration","source/dialogy.types.entity.keyword","source/dialogy.types.entity.numerical","source/dialogy.types.entity.people","source/dialogy.types.entity.time","source/dialogy.types.entity.time_interval","source/dialogy.types.intent","source/dialogy.types.plugin","source/dialogy.types.signal","source/dialogy.types.slots","source/dialogy.types.utterances","source/dialogy.utils","source/dialogy.workflow","source/modules"],envversion:{"sphinx.domains.c":2,"sphinx.domains.changeset":1,"sphinx.domains.citation":1,"sphinx.domains.cpp":4,"sphinx.domains.index":1,"sphinx.domains.javascript":2,"sphinx.domains.math":2,"sphinx.domains.python":3,"sphinx.domains.rst":2,"sphinx.domains.std":2,"sphinx.ext.viewcode":1,sphinx:56},filenames:["index.rst","source/dialogy.rst","source/dialogy.base.rst","source/dialogy.base.entity_extractor.rst","source/dialogy.base.input.rst","source/dialogy.base.output.rst","source/dialogy.base.plugin.rst","source/dialogy.cli.rst","source/dialogy.constants.rst","source/dialogy.plugins.rst","source/dialogy.plugins.text.rst","source/dialogy.plugins.text.calibration.rst","source/dialogy.plugins.text.canonicalization.rst","source/dialogy.plugins.text.classification.rst","source/dialogy.plugins.text.combine_date_time.rst","source/dialogy.plugins.text.duckling_plugin.rst","source/dialogy.plugins.text.lb_plugin.rst","source/dialogy.plugins.text.list_entity_plugin.rst","source/dialogy.plugins.text.list_search_plugin.rst","source/dialogy.plugins.text.merge_asr_output.rst","source/dialogy.plugins.text.slot_filler.rst","source/dialogy.plugins.text.voting.rst","source/dialogy.types.rst","source/dialogy.types.entity.rst","source/dialogy.types.entity.amount_of_money.rst","source/dialogy.types.entity.base_entity.rst","source/dialogy.types.entity.credit_card_number.rst","source/dialogy.types.entity.deserialize.rst","source/dialogy.types.entity.duration.rst","source/dialogy.types.entity.keyword.rst","source/dialogy.types.entity.numerical.rst","source/dialogy.types.entity.people.rst","source/dialogy.types.entity.time.rst","source/dialogy.types.entity.time_interval.rst","source/dialogy.types.intent.rst","source/dialogy.types.plugin.rst","source/dialogy.types.signal.rst","source/dialogy.types.slots.rst","source/dialogy.types.utterances.rst","source/dialogy.utils.rst","source/dialogy.workflow.rst","source/modules.rst"],objects:{"":[[1,0,0,"-","dialogy"]],"dialogy.base":[[3,0,0,"-","entity_extractor"],[6,0,0,"-","plugin"]],"dialogy.base.entity_extractor":[[3,1,1,"","EntityScoringMixin"],[3,4,1,"","entity_scoring"]],"dialogy.base.entity_extractor.EntityScoringMixin":[[3,2,1,"","DATETIME_OPERATION_ALIAS"],[3,2,1,"","FUTURE"],[3,2,1,"","PAST"],[3,3,1,"","aggregate_entities"],[3,3,1,"","apply_filters"],[3,3,1,"","entity_consensus"],[3,3,1,"","make_transform_values"],[3,3,1,"","remove_low_scoring_entities"],[3,2,1,"","threshold"]],"dialogy.base.plugin":[[6,1,1,"","Plugin"]],"dialogy.base.plugin.Plugin":[[6,3,1,"","train"],[6,3,1,"","transform"],[6,3,1,"","utility"]],"dialogy.cli":[[7,4,1,"","add_project_command_arguments"],[7,4,1,"","add_workflow_command_arguments"],[7,4,1,"","main"],[7,0,0,"-","project"],[7,4,1,"","project_command_parser"],[7,0,0,"-","workflow"]],"dialogy.cli.project":[[7,4,1,"","manage_project"],[7,4,1,"","project_cli"]],"dialogy.cli.workflow":[[7,4,1,"","get_workflow"],[7,4,1,"","train_workflow"],[7,4,1,"","workflow_cli"]],"dialogy.constants":[[8,1,1,"","EntityKeys"],[8,1,1,"","SIGNAL"]],"dialogy.constants.EntityKeys":[[8,2,1,"","ALTERNATIVE_INDEX"],[8,2,1,"","ALTERNATIVE_INDICES"],[8,2,1,"","BODY"],[8,2,1,"","DIM"],[8,2,1,"","DURATION"],[8,2,1,"","END"],[8,2,1,"","FROM"],[8,2,1,"","GRAIN"],[8,2,1,"","INTERVAL"],[8,2,1,"","LATENT"],[8,2,1,"","NORMALIZED"],[8,2,1,"","ORIGIN"],[8,2,1,"","RANGE"],[8,2,1,"","SLOT_NAMES"],[8,2,1,"","START"],[8,2,1,"","TO"],[8,2,1,"","TYPE"],[8,2,1,"","UNIT"],[8,2,1,"","VALUE"],[8,2,1,"","VALUES"]],"dialogy.constants.SIGNAL":[[8,2,1,"","NAME"],[8,2,1,"","REPRESENTATION"],[8,2,1,"","STRENGTH"]],"dialogy.plugins":[[10,0,0,"-","text"]],"dialogy.plugins.text":[[11,0,0,"-","calibration"],[12,0,0,"-","canonicalization"],[13,0,0,"-","classification"],[15,0,0,"-","duckling_plugin"],[17,0,0,"-","list_entity_plugin"],[19,0,0,"-","merge_asr_output"],[20,0,0,"-","slot_filler"],[21,0,0,"-","voting"]],"dialogy.plugins.text.calibration":[[11,0,0,"-","xgb"]],"dialogy.plugins.text.calibration.xgb":[[11,1,1,"","CalibrationModel"],[11,1,1,"","FeatureExtractor"]],"dialogy.plugins.text.calibration.xgb.CalibrationModel":[[11,3,1,"","filter_asr_output"],[11,3,1,"","inference"],[11,3,1,"","predict"],[11,3,1,"","save"],[11,3,1,"","train"],[11,3,1,"","transform"],[11,3,1,"","utility"],[11,3,1,"","validate"]],"dialogy.plugins.text.calibration.xgb.FeatureExtractor":[[11,3,1,"","features"],[11,3,1,"","fit"],[11,3,1,"","transform"]],"dialogy.plugins.text.canonicalization":[[12,1,1,"","CanonicalizationPlugin"],[12,4,1,"","get_entity_type"]],"dialogy.plugins.text.canonicalization.CanonicalizationPlugin":[[12,3,1,"","mask_transcript"],[12,3,1,"","transform"],[12,3,1,"","utility"]],"dialogy.plugins.text.classification":[[13,0,0,"-","xlmr"]],"dialogy.plugins.text.classification.xlmr":[[13,1,1,"","XLMRMultiClass"]],"dialogy.plugins.text.classification.xlmr.XLMRMultiClass":[[13,3,1,"","inference"],[13,3,1,"","init_model"],[13,3,1,"","load"],[13,3,1,"","save"],[13,3,1,"","train"],[13,3,1,"","utility"],[13,5,1,"","valid_labelencoder"],[13,3,1,"","validate"]],"dialogy.plugins.text.duckling_plugin":[[15,1,1,"","DucklingPlugin"]],"dialogy.plugins.text.duckling_plugin.DucklingPlugin":[[15,3,1,"","apply_filters"],[15,3,1,"","select_datetime"],[15,3,1,"","transform"],[15,3,1,"","utility"]],"dialogy.plugins.text.list_entity_plugin":[[17,1,1,"","ListEntityPlugin"]],"dialogy.plugins.text.list_entity_plugin.ListEntityPlugin":[[17,3,1,"","get_entities"],[17,3,1,"","ner_search"],[17,3,1,"","regex_search"],[17,3,1,"","transform"],[17,3,1,"","utility"]],"dialogy.plugins.text.merge_asr_output":[[19,1,1,"","MergeASROutputPlugin"],[19,4,1,"","merge_asr_output"]],"dialogy.plugins.text.merge_asr_output.MergeASROutputPlugin":[[19,3,1,"","transform"],[19,3,1,"","utility"]],"dialogy.plugins.text.slot_filler":[[20,0,0,"-","rule_slot_filler"]],"dialogy.plugins.text.slot_filler.rule_slot_filler":[[20,1,1,"","RuleBasedSlotFillerPlugin"]],"dialogy.plugins.text.slot_filler.rule_slot_filler.RuleBasedSlotFillerPlugin":[[20,3,1,"","fill"],[20,3,1,"","utility"]],"dialogy.types":[[23,0,0,"-","entity"],[34,0,0,"-","intent"],[35,0,0,"-","plugin"],[36,0,0,"-","signal"],[37,0,0,"-","slots"],[38,0,0,"-","utterances"]],"dialogy.types.entity":[[25,0,0,"-","base_entity"]],"dialogy.types.entity.base_entity":[[25,1,1,"","BaseEntity"],[25,4,1,"","entity_synthesis"]],"dialogy.types.entity.base_entity.BaseEntity":[[25,3,1,"","add_parser"],[25,3,1,"","copy"],[25,2,1,"","entity_type"],[25,3,1,"","from_dict"],[25,3,1,"","get_value"],[25,3,1,"","json"],[25,3,1,"","reshape"],[25,3,1,"","set_value"],[25,3,1,"","validate"],[25,2,1,"","value"]],"dialogy.types.intent":[[34,1,1,"","Intent"]],"dialogy.types.intent.Intent":[[34,3,1,"","add_parser"],[34,3,1,"","apply"],[34,3,1,"","cleanup"],[34,3,1,"","fill_slot"],[34,3,1,"","json"]],"dialogy.types.signal":[[36,0,0,"-","signal"]],"dialogy.types.slots":[[37,1,1,"","Slot"]],"dialogy.types.slots.Slot":[[37,3,1,"","add"],[37,3,1,"","clear"],[37,3,1,"","json"]],"dialogy.utils":[[39,0,0,"-","file_handler"],[39,0,0,"-","logger"],[39,0,0,"-","misc"],[39,0,0,"-","naive_lang_detect"],[39,0,0,"-","normalize_utterance"]],"dialogy.utils.file_handler":[[39,4,1,"","create_timestamps_path"],[39,4,1,"","load_file"],[39,4,1,"","save_file"]],"dialogy.utils.misc":[[39,4,1,"","dt2timestamp"],[39,4,1,"","traverse_dict"],[39,4,1,"","validate_type"]],"dialogy.utils.naive_lang_detect":[[39,4,1,"","lang_detect_from_text"]],"dialogy.utils.normalize_utterance":[[39,4,1,"","dict_get"],[39,4,1,"","is_each_element"],[39,4,1,"","is_list"],[39,4,1,"","is_list_of_string"],[39,4,1,"","is_string"],[39,4,1,"","is_unsqueezed_utterance"],[39,4,1,"","is_utterance"],[39,4,1,"","normalize"],[39,4,1,"","utterance2alternatives"]],"dialogy.workflow":[[40,0,0,"-","workflow"]],"dialogy.workflow.workflow":[[40,1,1,"","Workflow"]],"dialogy.workflow.workflow.Workflow":[[40,2,1,"","NON_SERIALIZABLE_FIELDS"],[40,3,1,"","execute"],[40,3,1,"","flush"],[40,2,1,"","input"],[40,3,1,"","json"],[40,2,1,"","lock"],[40,2,1,"","output"],[40,3,1,"","run"],[40,3,1,"","set_io"],[40,3,1,"","train"]],dialogy:[[2,0,0,"-","base"],[7,0,0,"-","cli"],[8,0,0,"-","constants"],[9,0,0,"-","plugins"],[22,0,0,"-","types"],[39,0,0,"-","utils"],[40,0,0,"-","workflow"]]},objnames:{"0":["py","module","Python module"],"1":["py","class","Python class"],"2":["py","attribute","Python attribute"],"3":["py","method","Python method"],"4":["py","function","Python function"],"5":["py","property","Python property"]},objtypes:{"0":"py:module","1":"py:class","2":"py:attribute","3":"py:method","4":"py:function","5":"py:property"},terms:{"0":[6,8,12,13,15,20,34,39],"02116":13,"1":[0,6,8,13,19,20,34,39],"10":[6,20],"11":[6,20,39],"12":[6,20],"13":20,"14":20,"15":20,"16":20,"1911":13,"2":[0,6,8,19,20,34,39],"3":[0,6,19,20,34,39],"30":[15,39],"4":[0,6,15,20,39],"42":15,"5":[6,13,15,20,39],"51":39,"52":[3,39],"53":39,"54":39,"55":39,"556a11057179":39,"6":[6,20,39],"639":15,"7":[6,15,20,39],"8":[6,20,34,39],"8000":15,"9":[6,20,39],"abstract":[0,6,40],"case":[6,13,15,19,34,39,40],"class":[0,3,6,8,11,12,13,15,17,19,20,25,34,37,40],"const":[25,39],"default":[7,13,15,25,39,40],"do":[15,40],"final":40,"float":[3,11,15],"function":[3,6,7,12,15,19,25,39,40],"import":[6,15,19,20,25,34,35,37,38,39],"int":[6,13,15,39],"long":[6,40],"new":[0,6,7,23,25],"return":[0,3,6,7,11,12,13,15,17,19,20,25,34,37,39,40],"static":3,"super":6,"true":[6,13,17,20,34,39,40],"try":[6,39],"while":25,A:[0,3,6,7,11,12,13,15,17,19,20,22,25,34,39,40],As:[6,20,40],If:[6,17,19,20,25,39,40],In:[6,13,15,19,20,34,39],It:[0,11,19,25,40],Its:25,Not:20,Of:15,One:17,Or:39,TO:8,That:19,The:[0,6,7,13,15,19,25,34,39,40],There:[6,15,17,22,34,40],These:[6,19,25,40],To:20,_:[6,20],__call__:15,__init__:6,__valid:25,_alternatives_:11,_error_:13,_thread:40,ab:13,abc:6,abl:[6,19],abort:7,about:[6,15],access:[6,11,12,13,15,17,19,20,23,25,35,39,40],accessor:6,acquir:3,across:15,act:22,action:[20,34,40],action_slot:20,activ:0,activate_latent_ent:15,ad:23,adapt:39,add:[6,25,37],add_pars:[25,34],add_project_command_argu:7,add_workflow_command_argu:7,addit:20,addition:6,address:[15,19],advanc:40,advantag:1,advent:15,affect:17,affirm:6,after:15,aggregate_ent:3,agnost:6,ahead:6,ai:[0,3,7,15,39],aim:40,air:22,all:[0,11,15,17,25,37,40],allocate_lock:40,allow:[6,17,20,40],along:0,alpha2:15,alreadi:[7,22],also:[0,6,15],altern:[6,11,12,15,17,19,38,39],alternative_index:[8,20,25,34],alternative_indic:[8,20,25],alwai:6,am:[11,20],am_scor:0,amount:0,amount_of_monei:[22,23],an:[0,6,7,11,13,15,19,20,22,25,34,39,40],ani:[0,3,6,7,11,12,13,15,17,19,20,23,25,34,37,39,40],anoth:[17,40],anticip:40,anyth:[6,40],apart:40,api:[0,15],appli:[6,15,25,34,39],applic:[0,15,40],apply_filt:[3,15],appropri:13,ar:[0,3,6,7,11,13,15,17,19,20,22,23,25,37,39,40],arbitrari:[25,39,40],arg:[6,7,11,12,13,15,17,19,20,40],args_map:13,argument:[0,6,7,39,40],argumentpars:7,arrai:6,arrang:20,artifact:[7,13,15,19,40],arxiv:13,asia:15,ask:19,asolitaryplugin:6,asr:[0,6,11,13,19,39],asr_output:11,assert:39,assist:[20,39],associ:[20,22,34],assum:[6,40],attribut:[15,25,34,40],attributeerror:13,audio:19,author:6,avail:[6,7,13,17,39],b:6,back:6,background:6,bad:[11,40],base:[0,1,8,11,12,13,15,17,19,20,25,34,37,40,41],base_ent:[22,23],baseent:[3,15,17,20,25,34,37,40],baseestim:11,becaus:[0,6,15,40],been:40,beep:39,befor:[0,11,39,40],behaviour:6,being:25,best:19,between:[6,11,20,22,40],bit:[6,11,20],bloat:25,block:[0,40],bodi:[3,8,20,25,34],bool:[6,11,13,17,20,34,39,40],bop:39,branch:[0,7],build:[0,6,7,40],built:[3,6],bunch:40,bundl:[17,40],c:6,calcul:13,calibr:[9,10],calibra:11,calibrationmodel:11,call:[6,11,12,13,15,17,19,20,25,34,39],callabl:[39,40],can:[0,3,6,13,15,17,19,20,22,34,37,40],candid:17,canonic:[9,10],canonicalizationplugin:12,card:0,care:40,categori:20,caution:[19,25],cd:[0,7],certain:[15,19,20,39],chanc:19,chang:[0,11,25],change_log_level:39,charact:39,check:[11,15,20,25,34,39,40],choic:[6,40],cl:25,classif:[6,9,10],classifi:[0,6,11,13,20,40],classmethod:25,cleanup:34,clear:37,clf:6,clf_featur:0,cli:[0,1,41],clone:0,code:[0,15,39],column:6,com:[0,3,15,20,39],combin:[0,3],combine_date_tim:[9,10],combinedatetimeoverslot:0,command:[0,7],command_str:7,commit:6,common:19,compar:15,compli:25,complianc:25,comput:40,concat_altern:6,concaten:[0,6,19],conceal:25,concept:40,conda:0,condit:40,condition:[3,15],conduit:40,confid:6,config:6,configur:[3,6,15,17,20,34,39],connect:[0,15],consid:[3,11,22],consider:40,consist:15,constant:[1,41],constrain:6,constraint:40,contain:[0,3,11,13,20,34],content:41,convent:6,convers:22,convert:[25,34,37,39],copi:[6,7,25],copier:7,core:[6,40],correspond:13,cost:19,could:40,count_peopl:20,countri:15,countvector:6,coupl:40,cover:[6,19],creat:[0,6,7,15,19,20,25,34,39],create_timestamps_path:39,creation:6,creator:6,credit:0,credit_card_numb:[22,23],csv:7,currenc:15,current:[0,6,17,20,37,40],cycl:6,d:[6,20],data:[3,6,7,11,12,13,15,17,19,25,40],data_column:13,databas:15,datafram:[6,11,12,13,15,17,19,40],datapoint:6,dataset:7,datastor:6,date:[0,15,22],date_tim:39,datetim:[1,15,41],datetime_filt:15,datetime_operation_alia:3,de:3,debug:[6,11,12,13,15,17,19,20,25,34,40],decisiontreeclassifi:6,deep:25,deepli:39,def:[0,6,15],defin:[6,15,20,25,35,39,40],definit:[19,34,37],demo:[6,34,40],depend:[0,3,6,23],deploi:15,deriv:[0,6,13],describ:[0,6,39],descript:[0,19,40],deseri:[22,23],design:[20,40],desir:40,destin:22,destination_path:7,detail:[19,34],detect:0,develop:[0,19],df:[6,11],dialog:20,dict:[3,11,15,17,19,25,34,37,39,40],dict_:25,dict_get:39,dict_travers:39,dictionari:[11,25,34,37,39],differ:[6,15,17,19,25,34],dim:[8,20,25],dimens:15,dir:7,directori:[0,7,39],discuss:6,dissimilar:19,distinct:6,dl:40,doc:0,document:19,doe:[6,20],doesn:[6,25,39],don:39,done:[6,19],down:15,download:[0,7],downstream:19,drop:40,dry:0,dry_run:39,dt2timestamp:39,dtype:6,duckl:[0,15,17,23],duckling_ent:15,duckling_pars:15,duckling_plugin:[9,10],ducklingplugin:[0,15,17,40],due:39,dump:6,duplic:3,durat:[0,8,15,22,23],dure:[6,11,12,13,17,19,20],dynam:6,e:6,each:[6,17,19,20,25,39],either:[6,22],element:[20,25,39,40],els:39,en_gb:15,en_in:15,en_u:15,encod:39,end:[6,8,19,20],end_token:6,enforc:40,english:20,ensur:[0,6],entiti:[0,1,3,6,12,15,17,20,22,34,37,39,40],entity_column:12,entity_consensu:3,entity_extractor:[1,2,15,17],entity_map:15,entity_scor:3,entity_synthesi:25,entity_typ:[20,25],entity_type_val_group:3,entity_type_value_group:3,entitykei:8,entityscoringmixin:[3,15,17],entri:7,env:39,ephemer:6,error:39,especi:15,etc:15,evalu:6,even:[3,20],everi:40,everyth:[6,39],exampl:[6,7,11,15,20,22],except:[6,39],exclus:39,execut:[6,40],exhibit:[7,40],exist:[0,6,7,39],exit:[0,7],expect:[6,7,15,17,19,20,22,25,39,40],experi:15,explicitli:23,explor:34,expos:[7,15],express:[0,15],extra:15,extract:[0,15,17,20,37],f:6,facilit:39,fact:19,fail:[6,39],fairli:[0,40],fallback_label:13,fals:[6,7,11,12,13,15,17,19,20,25,34,39,40],faq:20,far:6,fast:6,featur:[0,6,11],feature_extract:6,featureextractor:11,fed:6,few:[0,17,20,22,40],field:15,file:[6,7,39],file_handl:[1,41],file_nam:39,file_path:39,fill:[20,22,34,37,40],fill_multipl:[20,34],fill_slot:34,filler:[0,20],filter:[11,15,25],filter_asr_output:11,filter_typ:15,fine:39,first:11,fit:[6,11],flag:17,flatten:39,flavor:25,flexibl:40,flush:40,fn:7,fname:11,follow:[0,6,40],forc:0,foreign:40,fork:15,form:[19,39],format:[11,13,15,20,39,40],found:20,free:15,freedom:6,frequent:6,from:[3,6,7,8,11,13,15,17,19,20,25,34,39,40],from_dict:[17,25],fundament:0,futur:3,ge:3,gener:[6,7,22],geographi:15,get:[6,15,20,25,39,40],get_classifier_input:6,get_ent:17,get_entity_typ:12,get_valu:25,get_workflow:7,get_workflow_fn:7,git:[0,7],github:[0,3,7,15,39],gitlab:[0,7],given:[6,25,39],go:6,grain:8,graviti:40,great:19,greater:15,ground:40,group:3,guid:15,h:[0,7],ha:[0,6,25,34],hack:17,handl:[6,19,20,40],hang:40,happen:6,hard:0,have:[3,6,11,15,19,20,34,39,40],head:[0,7],hello:[6,7,39],helo:6,help:[0,7,15,19,25,34],henc:6,here:[7,15,39,40],hi:39,higher:[0,3],hold:34,hollow:40,hope:39,how:[0,6],howev:6,html:20,http:[3,13,15,20,39],i:[6,11,39],id:15,ideal:40,ident:[3,39],identifi:[20,25,34],imagin:6,impact:[17,19],imper:15,implement:[6,12,15,25],impli:[6,19],improv:19,inaud:11,includ:25,incompat:6,inconsequenti:19,index:39,indexerror:[25,39],individu:17,infer:[0,11,13,40],info:15,inform:[19,25,36],init_model:13,initi:[6,13],inner:40,input:[0,1,2,11,15,19,20,25,39,40],input_:[0,15,20,39,40],input_column:[6,11,12,15,17,19],input_s:3,insert:[17,37],instal:7,instanc:[6,7,15,25,34,37,39],instanti:[6,15,20,25],instead:[0,6,19,39],integ:39,intend:25,intent:[0,1,6,11,13,20,22,40],intent_detection_slot_fil:20,intent_nam:20,interact:[6,40],interfac:15,intermedi:17,interrupt:6,interv:8,intuit:6,invoc:[0,7],invok:15,involv:17,io:[17,40],ipython:39,irrespect:20,is_each_el:39,is_list:39,is_list_of_str:39,is_str:39,is_unsqueezed_utter:39,is_upd:7,is_utter:39,isn:[15,40],iso3166:15,iso:15,isol:6,issu:[3,15,23,39],item:[0,20],iter:40,its:[6,7,15,20,25,34],job:11,join:[6,19],json:[3,6,25,34,37,40],judg:40,just:[6,11,15,17,19,39],k:39,keep:[6,17,40],kei:[6,11,15,19,25,37,39],key_error:39,keyerror:[25,39],keyword:[22,23],keywordent:17,kind:6,know:[39,40],known:23,kolkata:15,kwarg:[7,13],label:[6,13,17],label_column:13,label_count:13,labelencod:13,lack:0,lambda:[15,20,39],lang_detect_from_text:39,languag:[15,39,40],largest:40,last:39,latent:[8,20,25],later:6,latest:[0,7],lb_plugin:[9,10],le:3,lead:39,learn:[11,40],least:13,leav:39,length:3,lesser:15,let:[6,20],level:40,li:[0,7],lib:39,librari:0,life:6,lifecycl:6,light:40,like:[6,15,19,20,22],line:[7,15],linear:6,link:7,lint:7,list:[0,3,6,11,12,13,15,17,19,20,25,37,39,40],list_entity_plugin:[9,10],list_search_plugin:[9,10],listentityplugin:17,littl:40,lm:[11,19],lm_score:0,load:[6,13,39,40],load_fil:39,load_model:40,loader:39,local:[7,15],locat:7,locationent:20,lock:40,log:40,logger:[1,41],logic:39,look:[6,20,39,40],lose:6,lower:[0,3],machin:40,mai:[3,6,19,20,39,40],main:7,make:[0,3,7,15,39,40],make_transform_valu:3,manag:[0,40],manage_project:7,mani:40,map:[17,20],mar:39,mark:19,mask:12,mask_token:12,mask_transcript:12,master:[0,7],mastercard:39,match:[6,15,17,20,25,39],matchtyp:17,mayb:34,maybe_utter:39,mean:[6,15,22,34],meant:25,medium:40,merg:19,merge_asr_output:[9,10],mergeasroutput:0,mergeasroutputplugin:19,mergeasrplugin:6,messag:[0,7],metadata:3,method:[0,11,12,13,15,17,19,20,25,40],migrat:0,millisecond:15,misc:[1,41],mismatch:39,miss:[7,19,39],mixin:3,mlp:[9,10],mock:40,mode:39,model:[0,6,11,13,17,19,20,39,40],model_dir:13,model_nam:11,modif:[25,34],modifi:[6,25,34,40],modul:[0,41],monei:0,more:[0,13,15,19,20,25,40],most:[3,19,39],multipl:[3,15,19,34],must:[6,7,39],mutat:[6,11,12,13,15,17,19,20,25,40],my:11,n:[13,19,39,40],naive_lang_detect:[1,41],name:[0,6,7,8,15,20,25,34,37,39,40],namespac:[0,7],natur:15,navig:39,neatli:40,necessari:25,need:[0,6,15,17,19,20,25,39],neg:6,neither:6,ner:17,ner_search:17,nest:39,newlin:39,next:[6,11,12,13,17,19,20],nlp:19,nlpprogress:20,nois:19,noisi:0,non:39,non_serializable_field:40,none:[3,6,7,11,12,13,15,17,19,20,25,34,39,40],nonetyp:7,nope:6,normal:[6,8,19,20,39],normalize_utter:[1,41],note:[6,15],noth:[25,40],notic:[15,19,20],now:6,number:[0,3,13,15,20,22],number_ent:20,number_slot:20,numer:[15,22,23],numericalent:20,o:39,obj:39,obj_typ:39,object:[3,6,7,8,15,25,34,37,39,40],obtain:15,off:6,offer:[6,40],onc:[6,15,20],one:[15,20,22,40],ones:[6,40],onli:[0,6,15,17,19,20,25,34],op:40,open:[6,23,39],oper:19,oppos:17,option:[0,3,6,7,13,15,17,20,25,39,40],order:0,org:13,organ:[0,7],origin:[8,20,22,25],other:[0,3,6,15,17,20,25,40],otherwis:13,our:6,out:[6,19,20,34,39],output:[0,1,2,11,15,19,20,39,40],output_column:[6,11,12,15,17,19],output_dir:7,over:[3,6,17,40],overal:15,overrid:[25,40],overridden:[25,40],overwrit:6,ow:19,own:[6,25],p1:6,p2:6,packag:41,panda:[6,13],param:[13,39],paramet:[3,7,11,13,15,17,19,20,25,34,39,40],pars:[6,15,17,25,39],parse_valu:6,parser:[7,15,20,25,34],part:15,partial:39,particip:22,pass:[0,6,11,12,13,15,17,19,20,40],passeng:22,past:3,path:[7,39,40],pattern:[15,17],pd:[6,11,13,15,17,40],peopl:[0,15,20,22,23],people_ent:20,peopleent:20,per:[6,15,20],perform:[6,19],pertain:20,phase:6,pick:25,pickl:6,piec:6,pip:0,pipe:40,pipelin:[6,11,40],piplin:11,pkl:11,place:[6,15],planet:39,plugin:[0,1,2,22,25,34,40,41],pluginfn:[6,17,20,35,40],poetri:[0,7],point:[3,6,7],portabl:40,posit:[0,7,19],possibl:[34,39],post:[6,40],postprocess:40,postprocessor:25,pprint:20,pr:0,practic:20,pre:[6,20,35,40],precis:19,predict:[6,11,13,40],prefer:[25,34,39],preprocess:40,preprocessor:15,presenc:3,present:20,preset:20,pretend:7,prevent:[15,17,25],previou:0,previous:6,pristin:17,privat:15,problem:[3,19],procedur:[6,40],process:[6,15,40],processor:40,produc:[0,6,15,19],product:[6,25,34],progress:[22,25,34],project:[1,8,22,40,41],project_cli:7,project_command_pars:7,prop:39,properti:[13,15,25,39],property_:25,proprietari:0,provid:[6,7,8,11,13,15,19,20,23,25,35,36,38,39],proxi:40,purpos:[6,7,13,25],push:6,put:40,py:39,pyenv:39,python3:39,python:[39,40],pytz:15,r:39,rais:[6,13,19,25,39,40],random_st:6,rang:[8,20,25,39],rank:0,rb:6,reach:15,read:[0,6,13,15],recal:19,receiv:[6,40],recent:39,recurs:39,redefin:6,reduc:[3,39],refactor:17,refer:[15,25,39],reference_time_column:15,reftim:15,regex:17,regex_search:17,regexflag:17,relai:40,relat:[15,20],relationship:[20,22],reli:6,remain:6,remedi:6,rememb:6,remov:[3,11,15,34,37],remove_low_scoring_ent:3,replac:25,report:20,repositori:0,repres:[19,20,40],represent:[8,25],requir:[0,6,15,17,22,40],reserv:22,reset:40,reshap:25,resid:39,respect:[19,22,25],respons:[0,15,39],restrict:40,result:[6,40],retain:0,roberta:13,role:6,root:[0,7],round:22,row:11,rtype:[25,39],rule:[20,34],rule_slot_fil:[9,10],rule_str:20,rulebasedslotfillerplugin:[20,34,40],run:[0,6,11,12,13,15,17,19,20,40],runtim:6,s:[0,6,7,15,17,19,20,34,39,40],safe:[25,39],safe_load:20,sai:40,same:[3,6,11,17,34,40],saniti:11,sauci:40,save:[6,11,13,39,40],save_fil:39,save_path:6,saw:6,scaffold:7,schema:20,score:[3,6,11,20,25,34],score_round_off:13,search:[3,15,17],searchstrategyclass:17,section:[0,6,15,22,25],see:[6,40],seemingli:6,select:15,select_datetim:15,self:[0,6],semant:19,send:6,sentenc:[15,19,22],separ:19,sequenc:6,serial:12,server:[0,15],servic:[15,17],set:[0,3,11,15,17,20,22,25,34],set_classifier_input:6,set_classifier_output:6,set_io:40,set_valu:25,setup:0,shall:20,shape:17,share:[3,34],sharp:11,shelf:6,shift:40,ship:[17,19,39],should:[0,6,11,19,25,34,40],show:[0,7,17],sic:40,signal:[1,8,22],signatur:[0,20,40],signific:[0,19,20,22],similar:[19,40],simpl:[0,7,19,39,40],simpli:19,sinc:[6,19,39],singl:[6,19,20,34],sit:11,site:39,skip:[13,25],skip_entity_attr:25,skip_label:13,skit:0,sklearn:[6,11],slightli:6,slot:[0,1,20,22,34,40],slot_fil:[9,10],slot_nam:[8,20,25],slot_rul:20,slot_typ:34,slow:15,slu:[0,11,22,40],smoothli:0,so:[6,17,19,20,22],some:[25,39,40],someon:6,somewher:15,somwher:6,sort:15,sound:6,sourc:[3,6,7,8,11,12,13,15,17,19,20,25,34,37,39,40],spaci:17,spacy_nlp:17,sparingli:39,speak:6,speaker:39,special:34,specif:[6,15,25,39],specifi:[6,7,15],stage:6,stall:15,standard:[15,39],start:[6,7,8,19,20],start_token:6,state:19,statist:40,still:6,str:[3,6,7,11,12,13,15,17,19,25,34,37,39,40],straight:6,strategi:[17,19],strength:8,string:[3,13,15,17,19,39],stronger:40,structur:[0,1,3,17,19,25],style:17,subclass:[7,15,25,40],submodul:[1,9,10,22,41],subpackag:41,sugar:[25,39],suitabl:40,supplier:40,support:[0,7,15,19,23,34,40],suppos:[25,40],sure:15,syntat:39,synthet:20,system:[19,20],t:[6,13,15,19,25,39,40],tag:[0,7,11],take:[6,7,40],task:[19,40],tell:[20,22],templat:[0,7],tend:[15,19],termin:19,test:6,text:[1,6,9,34,39],tfidf:[6,11],tfidftransform:6,than:[3,15,20],thei:[6,19,40],them:[19,20,40],therefor:6,thi:[0,3,6,7,11,12,13,15,17,19,20,22,25,34,36,37,39,40],think:40,third:20,those:[15,40],though:19,three:6,threshold:[3,11,12,13,15,17],through:[15,20,25,40],throughtout:8,thud:40,ticket:22,time:[0,6,15,17,20,22,23],time_interv:[22,23],time_slot:20,timeent:20,timeinterv:15,timeout:15,timestamp:39,timezon:15,todai:6,togeth:6,token:[9,10,20,22,25],too:0,tool:17,traceback:39,track:[0,15],train:[0,6,7,11,13,15,17,40],train_workflow:7,trainabl:[13,40],training_data:[6,11,12,13,15,17,19,40],transcript:[3,6,11,12,13,17,19,39],transcrit:11,transform:[7,11,12,13,15,17,19,20,39],transformermixin:11,trascript:11,travel:22,travers:39,traverse_dict:39,tree:6,treeclassifierplugin:6,trip:22,tupl:[3,11,15,20,39],turn:0,two:[11,20,40],type:[0,1,2,3,7,8,11,12,13,15,17,19,20,39,40,41],type_:39,typeerror:[19,39,40],tz:15,umm:39,under:19,understand:[0,6,34],union:[15,39],uniqu:[3,19],unit:[0,8,20],unix:39,unless:20,unsaf:25,unset:39,untrain:6,up:20,updat:[0,6,7,25,34,39,40],update_workflow:15,updated_altern:6,upon:22,url:15,us:[0,6,7,8,15,17,19,20,22,25,34,39,40],usag:[0,7],use_cuda:13,use_mast:7,use_transform:[6,11,12,15,17,19],user:[0,7],utf:39,util:[0,1,6,7,11,12,13,15,17,19,20,25,34,41],utter:[0,1,11,19,22,39],utterance2altern:39,valid:[6,11,13,15,25,40],valid_labelencod:13,validate_typ:39,valu:[3,6,8,11,12,13,15,17,19,20,25,34,37,39,40],valueerror:13,variou:39,vc:7,vcs_ref:7,vect:6,vector:[11,19],venv_dialog:39,veri:[19,40],verifi:25,vernacular:[3,7,15,39],version:[0,39,40],vertic:40,via:[6,17],virtualenv:0,visual:6,vote:[9,10],voteplugin:[19,40],w:[6,20,39],wa:[19,39],wai:[6,15,22],want:[6,11,15,34,39,40],wb:6,we:[0,3,6,7,11,15,17,19,20,22,23,34,39,40],well:[6,15,19,40],wercalibrationplugin:0,were:[15,25,34,40],when:[15,39],where:[0,3,7,15,19,25,34,40],which:[3,15,17,19,25,34,39,40],why:40,within:[3,6,15,17,22,25,34,37,39,40],without:[6,40],won:40,word:6,worfklow:6,work:[0,19,22,39,40],workflow:[0,1,6,11,12,13,15,17,19,20,41],workflow_cli:7,world:[6,7,39],worri:40,would:[6,15,19,22,40],wrap:39,wrapper:17,write:[0,2,39],writer:39,written:6,x:[6,11,12,13,17,19,20,39],x_col:6,xgb:[9,10],xgboostregressor:11,xlm:13,xlmr:[9,10],xlmrmulticlass:[0,13],y:[6,11,12,13,17,19,20,40],y_col:6,yaml:20,ye:[6,40],yea:6,you:[0,6,20,22,39,40],your:[0,6,40]},titles:["Dialogy","dialogy package","dialogy.base package","dialogy.base.entity_extractor package","dialogy.base.input package","dialogy.base.output package","dialogy.base.plugin package","dialogy.cli package","dialogy.constants package","dialogy.plugins package","dialogy.plugins.text package","dialogy.plugins.text.calibration package","dialogy.plugins.text.canonicalization package","dialogy.plugins.text.classification package","dialogy.plugins.text.combine_date_time package","dialogy.plugins.text.duckling_plugin package","dialogy.plugins.text.lb_plugin package","dialogy.plugins.text.list_entity_plugin package","dialogy.plugins.text.list_search_plugin package","dialogy.plugins.text.merge_asr_output package","dialogy.plugins.text.slot_filler package","dialogy.plugins.text.voting package","dialogy.types package","dialogy.types.entity package","dialogy.types.entity.amount_of_money package","dialogy.types.entity.base_entity package","dialogy.types.entity.credit_card_number package","dialogy.types.entity.deserialize package","dialogy.types.entity.duration package","dialogy.types.entity.keyword package","dialogy.types.entity.numerical package","dialogy.types.entity.people package","dialogy.types.entity.time package","dialogy.types.entity.time_interval package","dialogy.types.intent package","dialogy.types.plugin package","dialogy.types.signal package","dialogy.types.slots package","dialogy.types.utterances package","dialogy.utils package","dialogy.workflow package","dialogy"],titleterms:{advantag:40,amount_of_monei:24,base:[2,3,4,5,6],base_ent:25,calibr:11,canonic:12,classif:13,cli:7,combine_date_tim:14,constant:8,content:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],contributor:0,creation:0,credit_card_numb:26,datetim:39,deseri:27,dialogi:[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41],duckling_plugin:15,durat:28,entiti:[23,24,25,26,27,28,29,30,31,32,33],entity_extractor:3,file_handl:39,get:0,index:0,input:[4,6],instal:0,integr:6,intent:34,keyword:29,lb_plugin:16,list_entity_plugin:17,list_search_plugin:18,logger:39,merge_asr_output:19,method:6,misc:39,mlp:13,modul:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],naive_lang_detect:39,normalize_utter:39,numer:30,output:[5,6],packag:[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],peopl:31,plugin:[6,9,10,11,12,13,14,15,16,17,18,19,20,21,35],project:[0,7],rule_slot_fil:20,signal:36,slot:37,slot_fil:20,solitari:6,start:0,structur:40,submodul:[7,11,13,20,36,39,40],subpackag:[1,2,9,10,22,23],test:0,text:[10,11,12,13,14,15,16,17,18,19,20,21],time:32,time_interv:33,token:13,topic:0,trainabl:6,transform:6,type:[6,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],util:39,utter:38,vote:21,workflow:[7,40],write:6,xgb:11,xlmr:13}}) \ No newline at end of file diff --git a/docs/source/dialogy.base.entity_extractor.html b/docs/source/dialogy.base.entity_extractor.html index 6cc3fc31..debb8679 100644 --- a/docs/source/dialogy.base.entity_extractor.html +++ b/docs/source/dialogy.base.entity_extractor.html @@ -6,7 +6,7 @@ - dialogy.base.entity_extractor package - dialogy 0.9.3 documentation + dialogy.base.entity_extractor package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -226,20 +226,24 @@

    dialogy.base.entity_extractor package class EntityScoringMixin[source]

    Bases: object

    -

    Mixin for scoring and aggregation of entities over a set of transcripts.

    -
    +

    Mixin class for scoring entities acquired over a set of transcripts.

    +
    +
    +DATETIME_OPERATION_ALIAS = {'future': <built-in function ge>, 'past': <built-in function le>}
    +
    +
    +
    +FUTURE = 'future'
    +
    +
    +
    +PAST = 'past'
    +
    +
    aggregate_entities(entity_type_value_group, input_size)[source]

    Reduce entities sharing same type and value.

    -
      -

    • Entities with same type and value are considered identical even if other metadata is same. -These entities are part of a group.

      • -

    • We track the transcript indices for every entity in a group.

      • -

    • Select the minimum of all the indices. (because 0th transcript has highest confidence)

      • -

    • We pick one entity per group and modify its index to the minimum and score is aggregated for the group.

      • -

    • The entity picked is added to a list of aggregates.

      • -
    -

    The above is done for all entities in a group

    +

    Entities with same type and value are considered identical even if other metadata is same.

    Parameters

    entity_type_val_group (Dict[Tuple[str, Any], List[BaseEntity]]) – A data-structure that groups entities by type and value.

    @@ -255,7 +259,7 @@

    dialogy.base.entity_extractor package
    apply_filters(entities)[source]
    -

    Filter entities with score less than the threshold.

    +

    Conditionally remove entities.

    Parameters

    entities (List[BaseEntity]) – A list of entities.

    @@ -308,7 +312,6 @@

    dialogy.base.entity_extractor package remove_low_scoring_entities(entities)[source]

    Remove entities with a lower score than the threshold.

    -

    This doesn’t apply to entities with score=None.

    Parameters

    entities (List[BaseEntity]) – A list of entities.

    @@ -324,8 +327,7 @@

    dialogy.base.entity_extractor package
    threshold: Optional[float] = None
    -

    Value to compare against an entity’s score.

    -

    +

    diff --git a/docs/source/dialogy.base.html b/docs/source/dialogy.base.html index 199f02a6..d362f1f4 100644 --- a/docs/source/dialogy.base.html +++ b/docs/source/dialogy.base.html @@ -6,7 +6,7 @@ - dialogy.base package - dialogy 0.9.3 documentation + dialogy.base package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -229,34 +229,17 @@

    Subpackagesdialogy.base.input package
  • dialogy.base.output package
  • dialogy.base.plugin package

    • -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,241 +220,8 @@

    dialogy.base.input package

    -
    -

    Module contents

    -

    The Input class creates immutable instances that describe the inputs of a single turn of a conversation. -There are some attributes that may have aggregations of previous turns like the slot_tracker or entire history.

    -
    -

    Why do I see Input and Output as inputs to all Plugins?

    -

    It is a common pattern for all the plugins to require both as arguments. Since this could be confusing nomenclature, Input -and Output bear meaning and even separation for the SLU API, not for Plugins.

    -
    -
    -

    Updates

    -

    While writing plugins, we would need to update the attributes of Input, the following doesn’t work!

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.utils import make_unix_ts
-   ...: 
-
-In [2]: # Check the attributes in the object logged below. 
-   ...: input_x = Input(utterances="hello world", lang="hi", timezone="Asia/Kolkata")
-   ...: 
-
-In [3]: input_x
-Out[3]: Input(utterances='hello world', reference_time=None, latent_entities=False, transcripts=['hello world'], clf_feature=[], lang='hi', locale='en_IN', timezone='Asia/Kolkata', slot_tracker=None, current_state=None, previous_intent=None, history=None)
-
    -
    -
    -

    Issues with Frozen Instance Update

    -

    Now if we try the following:

    -
    In [4]: input_x.utterances = "hello"
----------------------------------------------------------------------------
-FrozenInstanceError                       Traceback (most recent call last)
-Input In [4], in <module>
-----> 1 input_x.utterances = "hello"
-
-File ~/miniconda3/envs/dialogy/lib/python3.9/site-packages/attr/_make.py:553, in _frozen_setattrs(self, name, value)
-    549 def _frozen_setattrs(self, name, value):
-    550     """
-    551     Attached to frozen classes as __setattr__.
-    552     """
---> 553     raise FrozenInstanceError()
-
-FrozenInstanceError: 
-
    -
    -

    We can see re-assigning values to attributes isn’t allowed.

    -
    -
    -

    Updating a frozen instance

    -

    We have to create new instances, but we have some syntax for it:

    -
    In [5]: input_x = Input(utterances="hello world", lang="hi", timezone="Asia/Kolkata")
-
-In [6]: input_y = Input.from_dict({"utterances": "hello"})
-
-In [7]: input_y
-Out[7]: Input(utterances='hello', reference_time=None, latent_entities=False, transcripts=['hello'], clf_feature=[], lang='en', locale='en_IN', timezone='UTC', slot_tracker=None, current_state=None, previous_intent=None, history=None)
-
    -
    -

    but by doing this, we lost the lang and timezone attributes to system defaults.

    -
    -
    -

    Reusing an existing instance

    -

    We can re-use an existing instance to create a new. -This way, we don’t have to write every existing property on a previous Input.

    -
    In [8]: input_x = Input(utterances="hello world", lang="hi", timezone="Asia/Kolkata")
-
-In [9]: input_y = Input.from_dict({"utterances": "hello"}, reference=input_x)
-
-In [10]: input_y
-Out[10]: Input(utterances='hello', reference_time=None, latent_entities=False, transcripts=['hello'], clf_feature=[], lang='hi', locale='en_IN', timezone='Asia/Kolkata', slot_tracker=None, current_state=None, previous_intent=None, history=None)
-
    -
    -
    -
    -
    -

    Serialization

    -

    If there is a need to represent an Input as a dict we can do the following:

    -
    In [1]: input_y.json()
-Out[1]: 
-{'utterances': 'hello',
- 'reference_time': None,
- 'latent_entities': False,
- 'transcripts': ['hello'],
- 'clf_feature': [],
- 'lang': 'hi',
- 'locale': 'en_IN',
- 'timezone': 'Asia/Kolkata',
- 'slot_tracker': None,
- 'current_state': None,
- 'previous_intent': None,
- 'history': None}
-
    -
    -
    -
    -class Input(transcripts=None, *, utterances, reference_time=None, latent_entities=False, clf_feature=NOTHING, lang='en', locale='en_IN', timezone='UTC', slot_tracker=None, current_state=None, previous_intent=None, history=None)[source]
    -

    Bases: object

    -

    Represents the inputs of the SLU API.

    -
    -
    -clf_feature: Optional[List[str]]
    -

    Placeholder for the features of an intent classifier.

    -
    ["<s> I want to book a flight </s> <s> I want to book flights </s> <s> I want to book a flight to Paris </s>"]
-
    -
    -
    -
    -
    -current_state: Optional[str]
    -

    Points at the active state (or node) within the conversation graph.

    -
    -
    -
    -classmethod from_dict(d, reference=None)[source]
    -

    Create a new Input instance from a dictionary.

    -
    -
    Parameters
    -
      -

    • d (Dict[str, Any]) – A dictionary such that keys are a subset of Input attributes.

      • -

    • reference (Optional[Input], optional) – An existing Input instance., defaults to None

      • -
    -
    -
    Returns
    -

    A new Input instance.

    -
    -
    Return type
    -

    Input

    -
    -
    -
    -
    -
    -history: Optional[List[Dict[str, Any]]]
    -
    -
    -
    -json()[source]
    -

    Serialize Input to a JSON-like dict.

    -
    -
    Returns
    -

    A dictionary that represents an Input instance.

    -
    -
    Return type
    -

    Dict[str, Any]

    -
    -
    -
    -
    -
    -lang: str
    -

    Expected language of the input. This is needed for language dependent plugins. -These are present in https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes -for “English” the code is “en”.

    -
    -
    -
    -latent_entities: bool
    -

    A switch to turn on/off production of latent entities via Duckling API -If you need to parse “4” as 4 am or 4 pm. Note the absence of “am” or “pm” in the utterance. -Then you might need this to be set as True. -It may be helpful to keep it False unless clearly required.

    -
    -
    -
    -locale: str
    -

    The locale identifier consists of at least a language code and a country/region code. -We keep “en_IN” as our default. This is used by Duckling for parsing patterns as per the -locale. If locale is missing i.e. None, we may fallback to lang instead.

    -
    -
    -
    -previous_intent: Optional[str]
    -

    The name of the intent that was predicted by the classifier in (exactly) the previous turn.

    -
    -
    -
    -reference_time: Optional[int]
    -

    The time that should be used for parsing relative time values. -This is a Unix timestamp in seconds. -utils/datetime.py has make_unix_ts to help convert -a date in ISO 8601 format to unix ms timestamp.

    -
    -
    -
    -slot_tracker: Optional[List[Dict[str, Any]]]
    -

    This data structure tracks the slots that were filled in previous turns. -This may come handy if we want to filter or reduce entities depending on our history. -We use this in our CombineDateTimeOverSlots plugin.

    -
    [{
-    "name": "_callback_",               # the intent name
-    "slots": [{
-        "name": "callback_datetime",    # the slot name
-        "type": [                       # can fill entities of these types.
-            "time",
-            "date",
-            "datetime"
-        ],
-        "values": [{                    # entities that were filled previously
-            "alternative_index": 0,
-            "body": "tomorrow",
-            "entity_type": "date",
-            "grain": "day",
-            "parsers": ["duckling"],
-            "range": {
-                "end": 8,
-                "start": 0
-            },
-            "score": None,
-            "type": "value",
-            "value": "2021-10-15T00:00:00+05:30"
-        }]
-    }]
-}]
-
    -
    -
    -
    -
    -timezone: str
    -

    Timezones from https://en.wikipedia.org/wiki/List_of_tz_database_time_zones -Used by duckling or any other date/time parsing plugins.

    -
    -
    -
    -transcripts: List[str]
    -

    A derived attribute. We cross product each utterance and return a list of strings. -We use this to normalize utterances.

    -
    -
    -
    -utterances: List[List[Dict[str, Optional[Union[str, float]]]]]
    -

    ASRs produce utterances. -Each utterance contains N-hypothesis. -Each hypothesis is a dict with keys "transcript" and "confidence".

    -
    -
    -
    +
    +

    Module contents

    @@ -510,16 +277,7 @@

    Serialization diff --git a/docs/source/dialogy.base.output.html b/docs/source/dialogy.base.output.html index 8f79c723..02d6adea 100644 --- a/docs/source/dialogy.base.output.html +++ b/docs/source/dialogy.base.output.html @@ -6,7 +6,7 @@ - dialogy.base.output package - dialogy 0.9.3 documentation + dialogy.base.output package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,105 +220,8 @@

    dialogy.base.output package

    -
    -

    Module contents

    -

    The Output class creates immutable instances that describe the output of a single turn of a conversation. -The output contains

    -
      -

    1. List[Intent] sorted in descending order of confidence scores.

      2. -

    2. List[BaseEntity] in an arbitrary order.

      3. -
    -

    We don’t require the List[BaseEntity] apart from logging purposes. We fill relevant entities within slots -of the predicted Intent. Currently this is done only for the Intent with the highest confidence score. -That’s because currently, we only process a single trigger from the SLU API (at the dialog manager), which is an Intent.

    -
    -

    Why do I see Input and Output as inputs to all Plugins?

    -

    It is a common pattern for all the plugins to require both as arguments. Since this could be confusing nomenclature, Input -and Output bear meaning and even separation for the SLU API, not for Plugins.

    -
    -
    -

    Setting Output

    -

    Just like Input, Output is also an immutable class.

    -
      -

    1. Don’t change the attribute elements in place. It would work for now -but even types like Intent and BaseEntity will soon become immutable as well.

      2. -

    2. Plugins that produce output must always produce a List[Intent] or List[BaseEntity]. -Say, returning a single Intent would break validations on the output class.

      3. -
    -
    -
    -

    Serialization

    -

    If there is a need to represent an Input as a dict we can do the following:

    -
    In [1]: from dialogy.base import Output
-   ...: from dialogy.types import Intent
-   ...: 
-
-In [2]: output = Output(intents=[Intent(name="_greeting_", score=1.0)])
-
-In [3]: output.json()
-Out[3]: 
-{'intents': [{'name': '_greeting_',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': []}
-
    -
    -
    -
    -class Output(*, intents=NOTHING, entities=NOTHING)[source]
    -

    Bases: object

    -

    Represents output of the SLU API.

    -
    -
    -entities: List[dialogy.types.entity.base_entity.BaseEntity]
    -

    A list of entities. Produced by DucklingPlugin, -ListEntityPlugin or ListSearchPlugin.

    -
    -
    -
    -classmethod from_dict(d, reference=None)[source]
    -

    Create a new Output instance from a dictionary.

    -
    -
    Parameters
    -
      -

    • d (Dict[str, Any]) – A dictionary such that keys are a subset of Output attributes.

      • -

    • reference (Optional[Output], optional) – An existing Output instance., defaults to None

      • -
    -
    -
    Returns
    -

    A new Output instance.

    -
    -
    Return type
    -

    Output

    -
    -
    -
    -
    -
    -intents: List[dialogy.types.intent.Intent]
    -

    A list of intents. Produced by XLMRMultiClass -or MLPMultiClass.

    -
    -
    -
    -json()[source]
    -

    Serialize Output to a JSON-like dict.

    -
    -
    Parameters
    -

    self (Output) – [description]

    -
    -
    Returns
    -

    [description]

    -
    -
    Return type
    -

    Dict[str, List[Dict[str, Any]]]

    -
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -374,11 +277,7 @@

    Serialization diff --git a/docs/source/dialogy.base.plugin.html b/docs/source/dialogy.base.plugin.html index dbd72a34..1154b1b8 100644 --- a/docs/source/dialogy.base.plugin.html +++ b/docs/source/dialogy.base.plugin.html @@ -6,7 +6,7 @@ - dialogy.base.plugin package - dialogy 0.9.3 documentation + dialogy.base.plugin package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,329 +222,331 @@

    dialogy.base.plugin package

    Module contents

    -
    -

    Abstract Plugin

    -

    A Plugin transforms the data within the workflow. We use plugins for producing intents, entities, -filling slots and even preparing intermediate data for other plugins. This package ships the abstract base class for creating -other plugins. These offer a base for a few already implemented features like book-keeping and Guards.

    -
    -

    Writing Plugins

    -
    -

    Attention

    -

    We will learn plugin creation by exploring some examples. It is highly advised to read -the Input and Output sections first.

    -
    -
    -

    Keyword Intent

    -

    Say we have to predict an intent _greeting_ if the first item in the transcripts have "hello". -This is assuming that the transcripts are ranked by their confidence scores in the descending order.

    -
    In [1]: from typing import List
+
    Writing Plugins
    
+
    Plugins are an abstraction over procedures that offer a specific functionality. A plugin abstracts out the way it is fed an input, and the way its output is pushed out.
+A discussion about plugins requires a bit of background about Worfklows, a workflow is an ephemeral datastore. This distinction of roles
+between a Plugin and a Workflow is a convention to ensure that neither the Plugin creation or its utility is constrained
+by their author’s choices.
    
+
    
+
    Plugin Types
    
+
    There are three types of plugins:
    
+
      
+
    1. Solitary
      2. 
+
    2. Transformer
      3. 
+
    3. Trainable
      4. 
+
    
+
    
+
    Solitary
    
+
    A plugin that only interacts during the runtime of a workflow is called a Base plugin.
+We can create a Base plugin like so:
    
+
    
+
    Methods
    
+
    The utility method describes the feature implemented by the plugin, this is the only required method for Base plugins. This method is called
+during a workflow’s run. If a plugin produces a value, it can also be updated within workflow. Any kind of file reading, config parsing, model loading
+should be done during the plugin’s initialization. utility methods must have fast execution times. This also means if there is any dynamic input required
+by a plugin, it should be available within the *args.
    
+
    
+
    In [1]: from typing import Any, Optional, List
+   ...: from dialogy.base.plugin import Plugin
+   ...: from dialogy.types import PluginFn
    ...: from dialogy.workflow import Workflow
-   ...: from dialogy.base import Input, Plugin, Output
-   ...: from dialogy.types import Intent
    ...: 
-
-In [2]: class Transcript2IntentPlugin(Plugin):
-   ...:     def __init__(self, dest=None, **kwargs):
-   ...:         super().__init__(dest=dest, **kwargs)
-   ...:         
-   ...:     def greet(self, transcripts: List[str]) -> List[Intent]:
-   ...:         if not transcripts:
-   ...:             return []
+   ...: class ASolitaryPlugin(Plugin):
+   ...:     def __init__(
+   ...:         self,
+   ...:         access: Optional[PluginFn] = None,
+   ...:         mutate: Optional[PluginFn] = None,
+   ...:         debug: bool = False
+   ...:     ):
+   ...:         super().__init__(access=access, mutate=mutate, debug=debug)
    ...: 
-   ...:         best_transcript, *rest = transcripts
+   ...:     def utility(self, *args: Any) -> Any:
+   ...:         return None
    ...: 
-   ...:         if "hello" in best_transcript:
-   ...:             return [Intent(name="_greeting_", score=1.0)]
-   ...:         else:
-   ...:             return []
+
    
+
    
+
    
+
    Let’s take an example of a plugin that concatenates the transcripts from an ASR to a str.
    
+
    
+
    In [2]: class MergeASRPlugin(Plugin):
+   ...:     def __init__(
+   ...:         self,
+   ...:         start_token: str = "<s>",
+   ...:         end_token: str = "</s>",
+   ...:         access: Optional[PluginFn] = None,
+   ...:         mutate: Optional[PluginFn] = None,
+   ...:         debug: bool = False
+   ...:     ):
+   ...:         super().__init__(access=access, mutate=mutate, debug=debug)
+   ...:         # Always validate the input.
+   ...:         self.start_token = start_token
+   ...:         self.end_token = start_token
    ...: 
-   ...:     def utility(self, input_: Input, output: Output) -> List[Intent]:
-   ...:         return self.greet(input_.transcripts)
+   ...:     def utility(self, *args: List[str]) -> Any:
+   ...:         # Always validate the input.
+   ...:         transcripts = args[0]
+   ...:         return self.start_token + f"{self.end_token} {self.start_token}".join(transcripts) + self.end_token
    ...: 
 
-In [3]: transcript2intent   = Transcript2IntentPlugin(dest="output.intents")
-   ...: workflow            = Workflow([transcript2intent])
-   ...: input_, output      = workflow.run(Input(utterances=[[{"transcript": "hello"}]]))
-   ...: # The last expression provides us a snapshot of the workflow's 
-   ...: # input and output after all plugins have been executed.
-   ...: # These aren't the Input and Output objects, but their `dict` equivalents.
+In [3]: MergeASRPlugin().utility(["hello world", "hello worlds", "hello word"])
+Out[3]: '<s>hello world<s> <s>hello worlds<s> <s>hello word<s>'
+
    
+
    
+
    
+
    
+
    
+
    Input
    
+
    Previously we saw that the utility method produces results on expected inputs. We will now see how to get the same from a workflow.
    
+
    
+
    In [4]: def get_classifier_input(workflow: Workflow) -> List[str]:
+   ...:     return (workflow.input["transcripts"],)
    ...: 
-
-In [4]: # we can see the utterance that we set 
-   ...: # and the derived field `transcript` that we used within the plugin.
-   ...: # Other values are defaults.
-   ...: input_
+
    
+
    
+
    
+
    This is an accessor function, these expect a workflow instance as their only argument and can return anything currently available.
+These accessor functions are written by workflow creators and allow someone to use core plugins as long as plugins specify the input, output types.
    
+
    
+
    
+
    Output
    
+
    Once the plugin produces an output, we may want to keep it somwhere in the workflow. In this case we rely on mutator functions to send
+the plugin output back to the workflow. These are also written by workflow creators and hence any updates are plugin agnostic.
    
+
    
+
    In [5]: def set_classifier_input(workflow: Workflow, value: str) -> None:
+   ...:     '''
+   ...:     We would normally overwrite the input but since a workflow
+   ...:     loses inputs once run; and returns only a copy of the output,
+   ...:     we are creating a new key within the output.
+   ...:     '''
+   ...:     workflow.output["transcripts"] = value
    ...: 
-Out[4]: 
-{'utterances': [[{'transcript': 'hello'}]],
- 'reference_time': None,
- 'latent_entities': False,
- 'transcripts': ['hello'],
- 'clf_feature': [],
- 'lang': 'en',
- 'locale': 'en_IN',
- 'timezone': 'UTC',
- 'slot_tracker': None,
- 'current_state': None,
- 'previous_intent': None,
- 'history': None}
-
-In [5]: # we can see that the _greeting_ intent has been set.
-   ...: output
+
    
+
    
+
    
+
    
+
    
+
    Integrations
    
+
    Now we fit the pieces together and can visualize the interactions between a workflow and a plugin.
    
+
    
+
    In [6]: w = Workflow([MergeASRPlugin(access=get_classifier_input, mutate=set_classifier_input)])
+   ...: w.output
+   ...: w.run({"transcripts": ["hello world", "hello worlds", "hello word"]})
    ...: 
-Out[5]: 
-{'intents': [{'name': '_greeting_',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': []}
+Out[6]: 
+{'intents': [],
+ 'entities': [],
+ 'transcripts': '<s>hello world<s> <s>hello worlds<s> <s>hello word<s>'}
 
    
 
    
+
    
 
    
-
    
-
    Heuristic based Intent
    
-
    This problem requires us to predict _greeting_ if the word "hello"
-is present at least 6 times across all utterances. To help us with this, 
-we will import an existing plugin: MergeASROutputPlugin.
    
-
    In [6]: from collections import Counter
-   ...: from dialogy.plugins import MergeASROutputPlugin
+
    
+
    
+
    Transformer
    
+
    Transformer plugins are solitary plugins that can additionally transform dataframes. These can be used for producing new features or modify existing ones.
+If we remember the MergeASRPlugin plugin we have built so far, it does perform its transformation on a single data point expected at runtime but transformations
+happen during the training phase of a workflow.
    
+
    Let’s try to build a transformer plugin that will add a new column to a dataframe.
    
+
    
+
    In [7]: import json
+   ...: import pandas as pd
    ...: 
-
-In [7]: class HeuristicBasedIntentPlugin(Plugin):
-   ...:     def __init__(self, threshold=0, dest=None, **kwargs):
-   ...:         super().__init__(dest=dest, **kwargs)
-   ...:         self.threshold = threshold
+   ...: class MergeASRPlugin(Plugin):
+   ...:     def __init__(
+   ...:         self,
+   ...:         start_token: str = "<s>",
+   ...:         end_token: str = "</s>",
+   ...:         access: Optional[PluginFn] = None,
+   ...:         mutate: Optional[PluginFn] = None,
+   ...:         input_column: str = "alternatives",
+   ...:         output_column: Optional[str] = "concat_alternatives",
+   ...:         use_transform: bool = True,
+   ...:         debug: bool = False
+   ...:     ):
+   ...:         super().__init__(access=access, mutate=mutate, input_column=input_column, output_column=output_column, use_transform=use_transform, debug=debug)
+   ...:         self.start_token = start_token
+   ...:         self.end_token = start_token
    ...: 
-   ...:     def greet(self, clf_features: List[str]) -> List[Intent]:
-   ...:         if not clf_features:
-   ...:             return []
+   ...:     def utility(self, *args: List[str]) -> Any:
+   ...:         transcripts = args[0]
+   ...:         return self.start_token + f"{self.end_token} {self.start_token}".join(transcripts) + self.end_token
    ...: 
-   ...:         feature = clf_features[0]
-   ...:         word_frequency = Counter(feature.split())
-   ...:         if word_frequency.get("hello", 0) >= self.threshold:
-   ...:             return [Intent(name="_greeting_", score=1.0)]
-   ...:         else:
-   ...:             return []
-   ...:     def utility(self, input_: Input, output: Output) -> List[Intent]:
-   ...:         return self.greet(input_.clf_feature)
+   ...:     def parse_value(self, value):
+   ...:         return self.utility(json.loads(value))
    ...: 
-
-In [8]: heuristic_based_intent  = HeuristicBasedIntentPlugin(threshold=6, dest="output.intents")
-   ...: merge_asr_output_plugin = MergeASROutputPlugin(dest="input.clf_feature")
-   ...: workflow                = Workflow([merge_asr_output_plugin, heuristic_based_intent])
-   ...: input_, output          = workflow.run(
-   ...:    Input(utterances=[
-   ...:        [{"transcript": "hello is anyone there"},
-   ...:         {"transcript": "yellow is anyone here"},
-   ...:         {"transcript": "hello is one here"},
-   ...:         {"transcript": "hello is one there"},
-   ...:         {"transcript": "hello if one here"},
-   ...:         {"transcript": "hello in one here"},
-   ...:         {"transcript": "hello ip one here"}]
-   ...:    ])
-   ...: )
+   ...:     def transform(self, training_data: pd.DataFrame):
+   ...:         if not self.use_transform:
+   ...:             return training_data
    ...: 
-
-In [9]: # let's check the snapshots again.
-   ...: # Pay close attention to `clf_feature`
-   ...: # it wasn't set in the previous example.
-   ...: # This time, it was set by the `MergeASROutputPlugin` plugin 
-   ...: # as it had the dest = "input.clf_feature"
-   ...: input_
+   ...:         training_data[self.output_column] = training_data[self.input_column].apply(self.parse_value)
+   ...:         return training_data
    ...: 
-Out[9]: 
-{'utterances': [[{'transcript': 'hello is anyone there'},
-   {'transcript': 'yellow is anyone here'},
-   {'transcript': 'hello is one here'},
-   {'transcript': 'hello is one there'},
-   {'transcript': 'hello if one here'},
-   {'transcript': 'hello in one here'},
-   {'transcript': 'hello ip one here'}]],
- 'reference_time': None,
- 'latent_entities': False,
- 'transcripts': ['hello is anyone there',
-  'yellow is anyone here',
-  'hello is one here',
-  'hello is one there',
-  'hello if one here',
-  'hello in one here',
-  'hello ip one here'],
- 'clf_feature': ['<s> hello is anyone there </s> <s> yellow is anyone here </s> <s> hello is one here </s> <s> hello is one there </s> <s> hello if one here </s> <s> hello in one here </s> <s> hello ip one here </s>'],
- 'lang': 'en',
- 'locale': 'en_IN',
- 'timezone': 'UTC',
- 'slot_tracker': None,
- 'current_state': None,
- 'previous_intent': None,
- 'history': None}
-
-In [10]: output
-Out[10]: 
-{'intents': [{'name': '_greeting_',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': []}
+   ...: df = pd.DataFrame({"alternatives": [json.dumps(["hello world", "helo world"]), json.dumps(["today only", "today"])]})
+   ...: MergeASRPlugin(input_column="alternatives", output_column="updated_alternatives").transform(df)
+   ...: 
+Out[7]: 
+                    alternatives                updated_alternatives
+0  ["hello world", "helo world"]  <s>hello world<s> <s>helo world<s>
+1        ["today only", "today"]        <s>today only<s> <s>today<s>
 
    
 
    
+
    
+
    and the lifecycle of its demo will be covered in a later section once we have an understanding of a trainable plugin.
    
 
    
-
    
-
    Guarding plugins
    
-
    It may seem useful to not run plugins all the time. In this case, if we knew the current state
-of the conversation we could decide to not run our naive plugin. Say a state where we are expecting numbers? 
-So let’s look at an example where we prevent plugins from execution using guards.
    
-
    In [11]: heuristic_based_intent  = HeuristicBasedIntentPlugin(
-   ....:     threshold=6,
-   ....:     dest="output.intents",
-   ....:     guards=[lambda i, o: i.current_state == "STATE_EXPECTING_NUMBERS"]
-   ....: )
-   ....: merge_asr_output_plugin = MergeASROutputPlugin(dest="input.clf_feature")
-   ....: workflow                = Workflow([merge_asr_output_plugin, heuristic_based_intent])
-   ....: in_                     = Input(utterances=[[
-   ....:                                 {"transcript": "hello is anyone there"},
-   ....:                                 {"transcript": "yellow is anyone here"},
-   ....:                                 {"transcript": "hello is one here"},
-   ....:                                 {"transcript": "hello is one there"},
-   ....:                                 {"transcript": "hello if one here"},
-   ....:                                 {"transcript": "hello in one here"},
-   ....:                                 {"transcript": "hello ip one here"}]]) 
-   ....: input_, output          = workflow.run(in_)
-   ....: 
-
-In [12]: # Oops we forgot to set the current state within the Input!
-   ....: output
-   ....: 
-Out[12]: 
-{'intents': [{'name': '_greeting_',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': []}
-
-In [13]: workflow            = Workflow([merge_asr_output_plugin, heuristic_based_intent])
-   ....: in_.current_state   = "STATE_EXPECTING_NUMBERS"
-   ....: # We get a FrozenInstanceError because Input and Output instances are immutable.
-   ....: 
----------------------------------------------------------------------------
-FrozenInstanceError                       Traceback (most recent call last)
-Input In [13], in <module>
-      1 workflow            = Workflow([merge_asr_output_plugin, heuristic_based_intent])
-----> 2 in_.current_state   = "STATE_EXPECTING_NUMBERS"
-
-File ~/miniconda3/envs/dialogy/lib/python3.9/site-packages/attr/_make.py:553, in _frozen_setattrs(self, name, value)
-    549 def _frozen_setattrs(self, name, value):
-    550     """
-    551     Attached to frozen classes as __setattr__.
-    552     """
---> 553     raise FrozenInstanceError()
+
    
+
    Trainable
    
+
    Trainable plugins are solitary plugins that can additionally train dataframes. A trainable plugin may optionally be able to transform dataframes as well.
+Now we will build a new demo plugin that will train a classifier on a dataframe.
    
+
    
+
    In [8]: import pickle
+   ...: from sklearn.tree import DecisionTreeClassifier
+   ...: from sklearn.pipeline import Pipeline
+   ...: from sklearn.feature_extraction.text import TfidfTransformer
+   ...: from sklearn.feature_extraction.text import CountVectorizer
+   ...: 
 
-FrozenInstanceError: 
+In [9]: class TreeClassifierPlugin(Plugin):
+   ...:     def __init__(
+   ...:         self,
+   ...:         access: Optional[PluginFn] = None,
+   ...:         mutate: Optional[PluginFn] = None,
+   ...:         random_state: int = 0,
+   ...:         x_col: str = "alternatives",
+   ...:         y_col: str = "intents",
+   ...:         debug: bool = False
+   ...:     ):
+   ...:         super().__init__(access=access, mutate=mutate, debug=debug)
+   ...:         self.model = Pipeline([
+   ...:             ('vect', CountVectorizer()),
+   ...:             ('tfidf', TfidfTransformer()),
+   ...:             ('clf', DecisionTreeClassifier(random_state=random_state))
+   ...:         ])
+   ...:         self.x_col = x_col
+   ...:         self.y_col = y_col
+   ...: 
+   ...:     def train(self, training_data: pd.DataFrame):
+   ...:         X = training_data[self.x_col]
+   ...:         y = training_data[self.y_col]
+   ...:         self.model.fit(X, y)
+   ...: 
+   ...:     def save(self):
+   ...:         with open(self.save_path, "wb") as handle:
+   ...:             pickle.dump(self.model, handle)
+   ...: 
+   ...:     def load(self):
+   ...:         with open(self.save_path, "rb") as handle:
+   ...:             self.model = pickle.load(handle)
+   ...: 
+   ...:     def utility(self, *args):
+   ...:         return self.model.predict(args)
+   ...: 
+   ...: df = pd.DataFrame({"alternatives": [json.dumps(["yes", "yea"]), json.dumps(["no", "nope"])], "intents": ["affirmative", "negative"]})
+   ...: 
+   ...: def set_classifier_output(workflow, value):
+   ...:    workflow.output["intents"] = value
+   ...: 
+   ...: def set_classifier_input(workflow, value):
+   ...:    workflow.input["transcripts"] = value
+   ...: 
 
-In [14]: # We can use the following to create a new instance of Input instead.
-   ....: in_ = Input.from_dict({"current_state": "STATE_EXPECTING_NUMBERS"}, reference=in_)
-   ....: input_, output  = workflow.run(in_)
-   ....: 
+In [10]: w = Workflow([MergeASRPlugin(access=get_classifier_input, mutate=set_classifier_input, input_column="alternatives", use_transform=True), TreeClassifierPlugin(access=get_classifier_input, mutate=set_classifier_output, x_col="concat_alternatives", y_col="intents")])
 
-In [15]: # In this case, we don't see anything set as the 
-   ....: # HeuristicBasedIntentPlugin was prevented
-   ....: # by its guarding conditions.
-   ....: output
-   ....: 
-Out[15]: {'intents': [], 'entities': []}
+In [11]: w.train(df)
 
-In [16]: # However, `clf_features` are set since we never wrote guards for 
-   ....: # the MergeASROutputPlugin.
-   ....: input_
-   ....: 
-Out[16]: 
-{'utterances': [[{'transcript': 'hello is anyone there'},
-   {'transcript': 'yellow is anyone here'},
-   {'transcript': 'hello is one here'},
-   {'transcript': 'hello is one there'},
-   {'transcript': 'hello if one here'},
-   {'transcript': 'hello in one here'},
-   {'transcript': 'hello ip one here'}]],
- 'reference_time': None,
- 'latent_entities': False,
- 'transcripts': ['hello is anyone there',
-  'yellow is anyone here',
-  'hello is one here',
-  'hello is one there',
-  'hello if one here',
-  'hello in one here',
-  'hello ip one here'],
- 'clf_feature': ['<s> hello is anyone there </s> <s> yellow is anyone here </s> <s> hello is one here </s> <s> hello is one there </s> <s> hello if one here </s> <s> hello in one here </s> <s> hello ip one here </s>'],
- 'lang': 'en',
- 'locale': 'en_IN',
- 'timezone': 'UTC',
- 'slot_tracker': None,
- 'current_state': 'STATE_EXPECTING_NUMBERS',
- 'previous_intent': None,
- 'history': None}
+In [12]: w.run({"transcripts": ["yes"]})
+Out[12]: {'intents': array(['affirmative'], dtype=object), 'entities': []}
 
    
 
    
-
    
-
    
-
    Update Plans
    
-
    You may need to write plugins that generate an Intent or BaseEntity, there
-is also a category of plugins that might be required for modifications. Like the RuleBasedSlotFillerPlugin
-or the CombineDateTimeOverSlots.
    
-
    The former updates the intents and the latter updates time-entities. In these cases, the plugins
-also take into account other values.
    
-
    Such as, CombineDateTimeOverSlots:
    
-
      
-
    1. Separates time entities from other entity types.
      2. 
-
    2. Combines them over slot presence.
      3. 
-
    3. Rebuilds a full entity list.
      4. 
-
    
-
    This means
    
-
      
-
    1. Some plugins need to replace the entire set of intents or entities.
      2. 
-
    2. While others need to append their results along with the history of the workflow.
      3. 
-
    
-
    If your plugin belongs to [1], then you need to set replace_output=True in the constructor of your plugin.
    
-
    
-
    Note
    
-
    Refer to CombineDateTimeOverSlots.
-It’s combine_time_entities_from_slots method shows where we may need to use
-replace_output=True.
    
-
    
+
    
+
    Notes:
    
+
      
+
    • If your plugin needs a model but it need not be trained frequently or is just an off the shelf pre-trained model, then you must go for a Solitary plugin.
      • 
+
    • A trainable plugin can also have transform methods if it must modify a dataframe for other plugins in place.
      • 
+
    
 
    
 
    
-class Plugin(input_column='alternatives', output_column=None, use_transform=False, replace_output=False, dest=None, guards=None, debug=False)[source]
    
+class Plugin(access=None, mutate=None, input_column='alternatives', output_column=None, use_transform=False, debug=False)[source]
 
    Bases: abc.ABC
    
-
    Abstract class to be implemented by all plugins.
    
-
    
-
    Parameters
    
-
      
-
    • input_column (str) – Transforms data in this column for a given dataframe, defaults to const.ALTERNATIVES
      • 
-
    • output_column (Optional[str]) – Saves transformation in this column for a given dataframe, defaults to None
      • 
-
    • use_transform (bool) – Should the transformation be applied while training?, defaults to False
      • 
-
    • dest (Optional[str]) – The path where plugin output should be saved., defaults to None
      • 
-
    • guards (Optional[List[Guard]]) – A list of functions that evaluate to bool, defaults to None
      • 
-
    • debug (bool, optional) – Should we print debug logs?, defaults to False
      • 
+
      A Plugin object provides utilities that perform a transformation. A Plugin object differs
+in behaviour depending on it’s purpose. Purpose defines stages in a Workflow’s lifecycle.
+These are namely: train, test and production.
      
+
      A Plugin object:
      
+
        
+
      • May be trained, expected during the train lifecycle.
        • 
+
      • May transform a dataframe, expected during the test lifecycle.
        • 
+
      • Transforms a datapoint, expected during the production lifecycle.
        • 
 
      
+
      Imagine a workflow processing a dataframe. Sequencing plugins like so:
      
+
      start -> A -> B -> C -> D -> end
+
      
+
      
+
      Let’s assume that the plugin B, and C are plugins that can transform data without any training.
+However, plugins A and D require training and transform data. The data transformations differ slightly as per
+the workflow life-cycle.
      
+
        
+
      1. 
+
        During the training phase:
          
+
        1. Plugin A starts training on a dataframe and then transforms the data.
          2. 
+
        2. Plugin B, C start transforming the dataframe received from A.
          3. 
+
        3. Plugin D starts trains on the dataframe received from C.
          4. 
+
        
 
        
 
        
-
        
-
        
-prevent(input_, output)[source]
        
-
        Decide if the plugin should execute.
        
-
        If this method returns true, the plugin’s utility method will not be called.
        
-
        
-
        Returns
        
-
        prevent plugin execution if True.
        
-
        
-
        Return type
        
-
        bool
        
+
        2. 
+
      2. 
+
        During the evaluation phase:
          
+
        1. Each plugin performs its own transformation and passes data ahead.
          2. 
+
        
 
        
 
        
-
    
+
+
+
    How does plugin P2 get trained if plugin P1 requires data from an incompatable dataframe?
    
+
    start -> P1 -> P2 -> end
+
    
+
    
+
    Given that the production lifecycle implies that a plugin either receives the base features
+straight from the Workflow or its derivates via transformations from pre-existing plugins.
    
+
    Therefore we can redefine incompatibility as follows:
    
+
      
+
    1. Label differences, i.e. P1 trains on Transcription dataframes but P2 trains on Intent Classification dataframes.
      2. 
+
    2. Feature differences, i.e. P1 trains on confidence scores but P2 trains on ASR transcripts.
      3. 
+
    
+
    Plugins require a method for validation to ensure the data they receive matches the data they are expected to process.
+If the validation fails, plugins must not interrupt the training by raising exceptions.
    
+
    Speaking from our example, P1 gets trained and P2 remaines untrained. This doesn’t sound intuitive but
+that’s because we are looking at plugins within a workflow. As a remedy, plugins
+can be trained in isolation. A plugin author still must provide train, transform and utility
+methods for a trainable plugin.
    
+
    Input
    
+
    Plugin instantiation provides the configuration but transformation is done by applying
+functions on a Workflow. This way the seemingly linear sequence of plugins has the freedom to
+access anything committed to the Workflow.
    
+
    Output
    
+
    Post transformation we need the plugin to commit a value. This value is not passed to the next plugin, instead we commit
+everything to the Workflow by applying a mutate function with the value over to the Workflow.
    
+
    Trainable Plugins
    
+
    If a Plugin is trainable, it must implement these methods:
    
+
      
+
    1. train(x: pd.DataFrame) -> Plugin
      2. 
+
    2. utility(...) -> Any
      3. 
+
    3. validate(x: pd.DataFrame) -> bool
      4. 
+
    4. save() -> Plugin
      5. 
+
    5. load() -> Plugin
      6. 
+
    
+
    Transformer Plugins
    
+
    If a Plugin transforms data and produces trainable features, it must implement these methods:
    
+
      
+
    1. transform(x: pd.DataFrame) -> pd.DataFrame
      2. 
+
    2. utility(...) -> Any
      3. 
+
    3. validate(x: pd.DataFrame) -> bool
      4. 
+
    
+
    Base Plugins
    
+
    If a Plugin is expected to transform isolated data points (no-training, no feature generation):
    
+
      
+
    1. utility(...) -> Any
      2. 
+
    
 
    
 
    
 train(_)[source]
    
@@ -567,20 +569,15 @@ 
    Update Plans
 
    
-abstract utility(input_, output)[source]
    
-
    An abstract method that describes the plugin’s functionality.
    
+abstract utility(*args)[source]
+
    Transform X -> y.
    
+
    This method is called during a Workflow run.
    
 
    
-
    Parameters
    
-
      
-
    • input (Input) – The workflow’s input.
      • 
-
    • output (Output) – The workflow’s output.
      • 
-
    
-
    
-
    Returns
    
-
    The value returned by the plugin.
    
+
    Returns
    
+
    A value that will be passed to the next plugin (if any).
    
 
    
-
    Return type
    
-
    Any
    
+
    Return type
    
+
    Any
    
 
    
 
    
 
    
@@ -588,6 +585,7 @@ 
    Update Plansdialogy.base.plugin package
    
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -284,7 +284,6 @@ 
    Submodules
 workflow_cli(args)[source]
 
    CLI entry point for workflows.
    
-
    args contain the following attributes:
    
 
    module      - We expect the Workflow or its subclass to be available at this path.
 fn          - A function that provides a Workflow or its subclass’ instance.
 data        - Path to dataset (.csv).
diff --git a/docs/source/dialogy.constants.html b/docs/source/dialogy.constants.html
index 075f7d59..2b25aa66 100644
--- a/docs/source/dialogy.constants.html
+++ b/docs/source/dialogy.constants.html
@@ -6,7 +6,7 @@
 
 
     
-        dialogy.constants package - dialogy 0.9.3 documentation
+        dialogy.constants package - dialogy 0.9.4 documentation
       
     
     
@@ -125,7 +125,7 @@
       
     
    
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -224,6 +224,91 @@ 
    dialogy.constants package
    Module contents
    
 
    Module provides Constants to be used throughtout the project.
    
 
    
+
    
+class EntityKeys[source]
    
+
    Bases: object
    
+
    
+
    
+ALTERNATIVE_INDEX = 'alternative_index'
    
+
    
+
    
+
    
+ALTERNATIVE_INDICES = 'alternative_indices'
    
+
    
+
    
+
    
+BODY = 'body'
    
+
    
+
    
+
    
+DIM = 'dim'
    
+
    
+
    
+
    
+DURATION = 'duration'
    
+
    
+
    
+
    
+END = 'end'
    
+
    
+
    
+
    
+FROM = 'from'
    
+
    
+
    
+
    
+GRAIN = 'grain'
    
+
    
+
    
+
    
+INTERVAL = 'interval'
    
+
    
+
    
+
    
+LATENT = 'latent'
    
+
    
+
    
+
    
+NORMALIZED = 'normalized'
    
+
    
+
    
+
    
+ORIGIN = 'origin'
    
+
    
+
    
+
    
+RANGE = 'range'
    
+
    
+
    
+
    
+SLOT_NAMES = 'slot_names'
    
+
    
+
    
+
    
+START = 'start'
    
+
    
+
    
+
    
+TO = 'to'
    
+
    
+
    
+
    
+TYPE = 'type'
    
+
    
+
    
+
    
+UNIT = 'unit'
    
+
    
+
    
+
    
+VALUE = 'value'
    
+
    
+
    
+
    
+VALUES = 'values'
    
+
    
+
    
+
    
 
    
 class SIGNAL[source]
    
 
    Bases: object
    
diff --git a/docs/source/dialogy.html b/docs/source/dialogy.html
index 86fdf97d..11a719bb 100644
--- a/docs/source/dialogy.html
+++ b/docs/source/dialogy.html
@@ -6,7 +6,7 @@
 
 
     
-        dialogy package - dialogy 0.9.3 documentation
+        dialogy package - dialogy 0.9.4 documentation
       
     
     
@@ -125,7 +125,7 @@
       
     
    
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -231,11 +231,11 @@ 
    Subpackagesdialogy.base.input package
 
 
  • dialogy.base.output package
 
    • 
 
  • dialogy.base.plugin package
    • 
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -238,14 +238,14 @@ 
    Subpackagesdialogy.plugins.text.classification package
 
 
  • dialogy.plugins.text.combine_date_time package
 
    • 
 
  • dialogy.plugins.text.duckling_plugin package
    • 
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -230,9 +230,9 @@ 
    Submodules
 
    
-class CalibrationModel(threshold, dest=None, guards=None, debug=False, input_column='alternatives', output_column='alternatives', use_transform=False, model_name='calibration.pkl')[source]
    
+class CalibrationModel(access, mutate, threshold, debug=False, input_column='alternatives', output_column='alternatives', use_transform=False, model_name='calibration.pkl')[source]
 
    Bases: dialogy.base.plugin.Plugin
    
-
    This plugin provides a calibration model that sits between ASR and SLU.
+
    This plugin provides a calibration model that sits between ASR and SLU.
 It trains a model that learn to classify alternatives from the text and
 AM, LM score. Bad alternatives are removed before training SLU and during
 inference.
    
@@ -255,7 +255,7 @@ 
    Submodules
 
    
-inference(transcripts, utterances)[source]
    
+inference(utterances)[source]
 
    
 
    Return type
    
 
    List[str]
    
@@ -308,20 +308,15 @@ 
    Submodules
 
    
-utility(input, _)[source]
    
-
    An abstract method that describes the plugin’s functionality.
    
+utility(*args)[source]
+
    Transform X -> y.
    
+
    This method is called during a Workflow run.
    
 
    
-
    Parameters
    
-
      
-
    • input (Input) – The workflow’s input.
      • 
-
    • output (Output) – The workflow’s output.
      • 
-
    
+
    Returns
    
+
    A value that will be passed to the next plugin (if any).
    
 
    
-
    Returns
    
-
    The value returned by the plugin.
    
-
    
-
    Return type
    
-
    Any
    
+
    Return type
    
+
    Any
    
 
    
 
    
 
    
diff --git a/docs/source/dialogy.plugins.text.canonicalization.html b/docs/source/dialogy.plugins.text.canonicalization.html
index 7c312e02..7aa72b78 100644
--- a/docs/source/dialogy.plugins.text.canonicalization.html
+++ b/docs/source/dialogy.plugins.text.canonicalization.html
@@ -6,7 +6,7 @@
 
 
     
-        dialogy.plugins.text.canonicalization package - dialogy 0.9.3 documentation
+        dialogy.plugins.text.canonicalization package - dialogy 0.9.4 documentation
       
     
     
@@ -125,7 +125,7 @@
       
     
    
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -224,7 +224,7 @@ 
    dialogy.plugins.text.canonicalization package
    Module contents
    
 
    
 
    
-class CanonicalizationPlugin(serializer=<function get_entity_type>, mask='MASK', mask_tokens=None, dest=None, guards=None, entity_column='entities', input_column='alternatives', output_column=None, use_transform=False, threshold=0.0, debug=False)[source]
    
+class CanonicalizationPlugin(serializer=<function get_entity_type>, mask='MASK', mask_tokens=None, access=None, mutate=None, entity_column='entities', input_column='alternatives', output_column=None, use_transform=False, threshold=0.0, debug=False)[source]
 
    Bases: dialogy.base.plugin.Plugin
    
 
    This plugin implements the canonicalization of the text.
    
 
    
@@ -248,20 +248,15 @@ 
    dialogy.plugins.text.canonicalization package
 
    
-utility(input, output)[source]
    
-
    An abstract method that describes the plugin’s functionality.
    
+utility(*args)[source]
+
    Transform X -> y.
    
+
    This method is called during a Workflow run.
    
 
    
-
    Parameters
    
-
      
-
    • input (Input) – The workflow’s input.
      • 
-
    • output (Output) – The workflow’s output.
      • 
-
    
-
    
-
    Returns
    
-
    The value returned by the plugin.
    
+
    Returns
    
+
    A value that will be passed to the next plugin (if any).
    
 
    
-
    Return type
    
-
    Any
    
+
    Return type
    
+
    Any
    
 
    
 
    
 
    
diff --git a/docs/source/dialogy.plugins.text.classification.html b/docs/source/dialogy.plugins.text.classification.html
index 2fe296fa..4bf67c49 100644
--- a/docs/source/dialogy.plugins.text.classification.html
+++ b/docs/source/dialogy.plugins.text.classification.html
@@ -6,7 +6,7 @@
 
 
     
-        dialogy.plugins.text.classification package - dialogy 0.9.3 documentation
+        dialogy.plugins.text.classification package - dialogy 0.9.4 documentation
       
     
     
@@ -125,7 +125,7 @@
       
     
    
     
    
-      
    dialogy 0.9.3 documentation
    
+      
    dialogy 0.9.4 documentation
    
     
    
     
    
       
    
@@ -148,7 +148,7 @@
       
    
   
   
-  dialogy 0.9.3 documentation
+  dialogy 0.9.4 documentation
   
 
   
@@ -223,167 +223,11 @@ 
    dialogy.plugins.text.classification package
 
    Submodules
    
 
    
-
    
-
    dialogy.plugins.text.classification.mlp module
    
-
    This module provides a trainable MLP classifier.
    
-
    
-
    
-class MLPMultiClass(model_dir, dest=None, guards=None, debug=False, threshold=0.1, score_round_off=5, purpose='train', fallback_label='_error_', data_column='data', label_column='labels', args_map=None, skip_labels=None, kwargs=None)[source]
    
-
    Bases: dialogy.base.plugin.Plugin
    
-
    This plugin provides a classifier based on sklearn’s MLPClassifier.
    
-
    
-
    
-static get_formatted_gridparams(params)[source]
    
-
    Gets the valid parameters for the gridsearch.
    
-
    
-
    Args:
    values: The values to be validated.
    
-
    
-
    Returns:
    The valid parameters.
    
-
    
-
    
-
    
-
    Return type
    
-
    List[Any]
    
-
    
-
    
-
    
-
    
-
    
-get_gridsearch_grid(**kwargs)[source]
    
-
    Gets gridsearch hyperparameters for the model in proper grid params format.
    
-
    
-
    Raises:
    ValueError: If a gridsearch parameter doesn’t exist in sklearns TFIDF and MLPClassifier modules.
    
-
    
-
    
-
    
-
    Return type
    
-
    List[Dict[str, List[Any]]]
    
-
    
-
    
-
    
-
    
-
    
-inference(texts)[source]
    
-
    Predict the intent of a list of texts.
    
-
    
-
    Parameters
    
-
    texts (List[str]) – A list of strings, derived from ASR transcripts.
    
-
    
-
    Raises
    
-
    AttributeError – In case the model isn’t of sklearn pipeline or gridsearchcv.
    
-
    
-
    Returns
    
-
    A list of intents corresponding to texts.
    
-
    
-
    Return type
    
-
    List[Intent]
    
-
    
-
    
-
    
-
    
-
    
-init_model()[source]
    
-
    Initialize the model if artifacts are available.
    
-
    
-
    Return type
    
-
    Dict[str, Any]
    
-
    
-
    
-
    
-
    
-
    
-load()[source]
    
-
    Load the plugin artifacts.
    
-
    
-
    Return type
    
-
    None
    
-
    
-
    
-
    
-
    
-
    
-save()[source]
    
-
    Save the plugin artifacts.
    
-
    
-
    Raises
    
-
    ValueError – In case the mlp model is not trained.
    
-
    
-
    Return type
    
-
    None
    
-
    
-
    
-
    
-
    
-
    
-train(training_data)[source]
    
-
    Train an intent-classifier on the provided training data.
    
-
    The training is skipped if the data-format is not valid.
-:type training_data: DataFrame
-:param training_data: A pandas dataframe containing at least list of strings and corresponding labels.
-:type training_data: pd.DataFrame
    
-
    
-
    Return type
    
-
    None
    
-
    
-
    
-
    
-
    
-
    
-utility(input, _)[source]
    
-
    An abstract method that describes the plugin’s functionality.
    
-
    
-
    Parameters
    
-
      
-
    • input (Input) – The workflow’s input.
      • 
-
    • output (Output) – The workflow’s output.
      • 
-
    
-
    
-
    Returns
    
-
    The value returned by the plugin.
    
-
    
-
    Return type
    
-
    Any
    
-
    
-
    
-
    
-
    
-
    
-property valid_mlpmodel: bool
    
-
    
-
    Return type
    
-
    bool
    
-
    
-
    
-
    
-
    
-
    
-validate(training_data)[source]
    
-
    Validate the training data is in the appropriate format
    
-
    
-
    Parameters
    
-
    training_data (pd.DataFrame) – A pandas dataframe containing at least list of strings and corresponding labels.
    
-
    
-
    Returns
    
-
    True if the dataframe is valid, False otherwise.
    
-
    
-
    Return type
    
-
    bool
    
-
    
-
    
-
    
-
    
+
    
+
    dialogy.plugins.text.classification.mlp module
    
 
    
-
    
-
    dialogy.plugins.text.classification.tokenizers module
    
-
    
-
    
-identity_tokenizer(text)[source]
    
-
    
-
    Return type
    
-
    str
    
-
    
-
    
-
    
+
    
+
    dialogy.plugins.text.classification.tokenizers module
    
 
    
 
    
 
    dialogy.plugins.text.classification.xlmr module
    
@@ -391,15 +235,13 @@ 
    Submoduleshttps://arxiv.org/abs/1911.02116)
    
 
    
 
    
-class XLMRMultiClass(model_dir, dest=None, guards=None, debug=False, threshold=0.1, use_cuda=False, score_round_off=5, purpose='train', fallback_label='_error_', use_state=False, data_column='data', label_column='labels', state_column='state', args_map=None, skip_labels=None, kwargs=None)[source]
    
+class XLMRMultiClass(model_dir, access=None, mutate=None, debug=False, threshold=0.1, use_cuda=False, score_round_off=5, purpose='train', fallback_label='_error_', data_column='data', label_column='labels', args_map=None, skip_labels=None, kwargs=None)[source]
 
    Bases: dialogy.base.plugin.Plugin
    
 
    This plugin provides a classifier based on XLM-Roberta <https://arxiv.org/abs/1911.02116>.
    
-
    The use_state flag in the XLMRMultiClass plugin is used to enable the use of state variable as the part of the text input.
    
 
    
 
    
 inference(texts)[source]
    
-
    Predict the intent of a list of texts.
-If the model has been trained using the state features, it expects the text to also be appended with the state token else the predictions would be spurious.
    
+
    Predict the intent of a list of texts.
    
 
    
 
    Parameters
    
 
    texts (List[str]) – A list of strings, derived from ASR transcripts.
    
@@ -459,7 +301,6 @@ 
    Submodulestrain(training_data)[source]
 
    Train an intent-classifier on the provided training data.
    
 
    The training is skipped if the data-format is not valid.
-While training with the use_state flag as true, make sure that the state column is the part of the training_data dataframe
 :type training_data: DataFrame
 :param training_data: A pandas dataframe containing at least list of strings and corresponding labels.
 :type training_data: pd.DataFrame
    
@@ -471,20 +312,15 @@ 
    Submodules
 
    
-utility(input, _)[source]
    
-
    An abstract method that describes the plugin’s functionality.
    
+utility(*args)[source]
+
    Transform X -> y.
    
+
    This method is called during a Workflow run.
    
 
    
-
    Parameters
    
-
      
-
    • input (Input) – The workflow’s input.
      • 
-
    • output (Output) – The workflow’s output.
      • 
-
    
-
    
-
    Returns
    
-
    The value returned by the plugin.
    
+
    Returns
    
+
    A value that will be passed to the next plugin (if any).
    
 
    
-
    Return type
    
-
    Any
    
+
    Return type
    
+
    Any
    
 
    
 
    
 
    
@@ -503,8 +339,7 @@ 
    Submodules
 
    Parameters
    
-
    training_data (pd.DataFrame) – A pandas dataframe containing at least list of strings and corresponding labels.
-Should also contain a state key if use_state = True while initializing object.
    
+
    training_data (pd.DataFrame) – A pandas dataframe containing at least list of strings and corresponding labels.
    
 
    
 
    Returns
    
 
    True if the dataframe is valid, False otherwise.
    
@@ -574,8 +409,8 @@ 
    Submodulesdialogy.plugins.text.classification package
diff --git a/docs/source/dialogy.plugins.text.combine_date_time.html b/docs/source/dialogy.plugins.text.combine_date_time.html
index 0a8140d5..95fe1fc8 100644
--- a/docs/source/dialogy.plugins.text.combine_date_time.html
+++ b/docs/source/dialogy.plugins.text.combine_date_time.html
@@ -6,7 +6,7 @@
 
 
     
-        dialogy.plugins.text.combine_date_time package - dialogy 0.9.3 documentation
+        dialogy.plugins.text.combine_date_time package - dialogy 0.9.4 documentation
       
     
     
@@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,175 +220,8 @@

    dialogy.plugins.text.combine_date_time package

    -
    -

    Module contents

    -
    -
    -class CombineDateTimeOverSlots(dest=None, guards=None, input_column='alternatives', output_column=None, replace_output=True, use_transform=False, trigger_intents=None, debug=False)[source]
    -

    Bases: dialogy.base.plugin.Plugin

    -

    Dialog case:

    -

    Assume that at the moment of dialog the date is 15th December 2019.

    -
    ------ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

    utterance

    entity

    1

    BOT

    When do you want to visit?

    2

    USER

    Day after tomorrow

    {‘type’: ‘date’, ‘value’: ‘2019-12-17’}

    3

    BOT

    At what time?

    4

    USER

    3 pm

    {‘type’: ‘time’, ‘value’: ‘2019-12-15T15:00:00+0000’}

    -

    We want to use the slots such that the date from interaction [2] and the time from interaction [4] are combined into a single entity.

    -

    We receive this as a tracked slot:

    -
    [{
-    "name": "_callback_",
-    "slots": [{
-        "name": "callback_datetime",
-        "type": [
-            "time",
-            "date",
-            "datetime"
-        ],
-        "values": [{
-            "alternative_index": 0,
-            "body": "tomorrow",
-            "entity_type": "date",
-            "grain": "day",
-            "parsers": ["duckling"],
-            "range": {
-                "end": 8,
-                "start": 0
-            },
-            "score": null,
-            "type": "value",
-            "value": "2021-10-15T00:00:00+05:30"
-        }]
-    }]
-}]
-
    -
    -

    At interaction [4] we would have a TimeEntity from DucklingPlugin but as we can see, the tracked slot provides us -previously filled values as json.

    -

    So we need to parse the first value in the slot (since they are the entities) if they are one of these date, time, datetime types. -This would give us a date / time entity that we can subsequently combine with the available entity in the current turn.

    -
    -
    -DATE = 'date'
    -
    -
    -
    -DATETIME = 'datetime'
    -
    -
    -
    -SUPPORTED_ENTITIES = ['date', 'time', 'datetime']
    -
    -
    -
    -TIME = 'time'
    -
    -
    -
    -combine_time_entities_from_slots(slot_tracker, entities)[source]
    -

    Combines the time entities from the slots with the entities from the current turn.

    -
    -
    Parameters
    -
      -

    • slot_tracker (Optional[List[Dict[str, Any]]]) – The slot tracker from the previous turn.

      • -

    • entities (List[BaseEntity]) – Entities found in the current turn.

      • -
    -
    -
    Returns
    -

    Combined set of entities.

    -
    -
    Return type
    -

    List[BaseEntity]

    -
    -
    -
    -
    -
    -get_tracked_slots(slot_tracker)[source]
    -
    -
    Return type
    -

    List[Dict[str, Any]]

    -
    -
    -
    -
    -
    -join(current_entity, previous_entity)[source]
    -
    -
    Return type
    -

    BaseEntity

    -
    -
    -
    -
    -
    -pick_previously_filled_time_entity(tracked_slots)[source]
    -
    -
    Return type
    -

    Optional[TimeEntity]

    -
    -
    -
    -
    -
    -utility(input, output)[source]
    -

    Combine the date and time entities collected across turns into a single entity.

    -
    -
    Return type
    -

    List[BaseEntity]

    -
    -
    -
    -
    -
    -
    -has_time_component(entity)[source]
    -
    -
    Return type
    -

    bool

    -
    -
    -
    -
    -
    -is_date(entity)[source]
    -
    -
    Return type
    -

    bool

    -
    -
    -
    +
    +

    Module contents

    @@ -444,7 +277,7 @@

    dialogy.plugins.text.combine_date_time package diff --git a/docs/source/dialogy.plugins.text.duckling_plugin.html b/docs/source/dialogy.plugins.text.duckling_plugin.html index fa37d03b..d31ba64b 100644 --- a/docs/source/dialogy.plugins.text.duckling_plugin.html +++ b/docs/source/dialogy.plugins.text.duckling_plugin.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.duckling_plugin package - dialogy 0.9.3 documentation + dialogy.plugins.text.duckling_plugin package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,713 +222,53 @@

    dialogy.plugins.text.duckling_plugin package

    Module contents

    -

    We use Duckling for parsing values like: date, time, -numbers, currency etc from natural language. The parser will expect Duckling to be running as an http service, and provide -means to connect from the implementation here. Here’s a big example for various duckling entities.

    -
    -

    Mother of all examples

    -
    In [1]: from dialogy.workflow import Workflow
-   ...: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=[
-   ...:         "number",
-   ...:         "people",
-   ...:         "time",
-   ...:         "duration",
-   ...:         "intersect",
-   ...:         "amount-of-money",
-   ...:         "credit-card-number",
-   ...:     ],
-   ...:     # Duckling supports multiple dimensions, by specifying a list, we make sure
-   ...:     # we are searching for a match within the expected dimensions only.
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: utterances = [
-   ...:    "between today 7pm and tomorrow 9pm",
-   ...:    "can we get up at 6 am",
-   ...:    "we are 7 people",
-   ...:    "2 children 1 man 3 girls",
-   ...:    "can I come now?",
-   ...:    "can I come tomorrow",
-   ...:    "how about monday then",
-   ...:    "call me on 5th march",
-   ...:    "I want 4 pizzas",
-   ...:    "2 hours",
-   ...:    "can I pay $5 instead?",
-   ...:    "my credit card number is 4111111111111111",
-   ...: ]
-   ...: 
-
-In [5]: %%time
-   ...: input_, output = workflow.run(Input(utterances=utterances))
-   ...: 
-CPU times: user 26.2 ms, sys: 2.56 ms, total: 28.7 ms
-Wall time: 30.8 ms
-
-In [6]: output
-Out[6]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 34},
-   'body': 'between today 7pm and tomorrow 9pm',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 0,
-   'entity_type': 'datetime',
-   'grain': 'hour',
-   'type': 'interval',
-   'from_value': '2022-02-15T19:00:00.000+00:00',
-   'to_value': '2022-02-16T22:00:00.000+00:00',
-   'value': {'from': '2022-02-15T19:00:00.000+00:00',
-    'to': '2022-02-16T22:00:00.000+00:00'}},
-  {'range': {'start': 14, 'end': 21},
-   'body': 'at 6 am',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 1,
-   'value': '2022-02-16T06:00:00.000+00:00',
-   'entity_type': 'time',
-   'grain': 'hour'},
-  {'range': {'start': 7, 'end': 15},
-   'body': '7 people',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 2,
-   'value': 7,
-   'unit': 'person',
-   'entity_type': 'people'},
-  {'range': {'start': 0, 'end': 10},
-   'body': '2 children',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 3,
-   'value': 2,
-   'unit': 'child',
-   'entity_type': 'people'},
-  {'range': {'start': 11, 'end': 16},
-   'body': '1 man',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 3,
-   'value': 1,
-   'unit': 'male',
-   'entity_type': 'people'},
-  {'range': {'start': 17, 'end': 24},
-   'body': '3 girls',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 3,
-   'value': 3,
-   'unit': 'female',
-   'entity_type': 'people'},
-  {'range': {'start': 11, 'end': 14},
-   'body': 'now',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 4,
-   'value': '2022-02-15T13:18:37.795+00:00',
-   'entity_type': 'datetime',
-   'grain': 'second'},
-  {'range': {'start': 11, 'end': 19},
-   'body': 'tomorrow',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 5,
-   'value': '2022-02-16T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 4, 'end': 16},
-   'body': 'about monday',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 6,
-   'value': '2022-02-21T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 8, 'end': 20},
-   'body': 'on 5th march',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 7,
-   'value': '2022-03-05T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 7, 'end': 8},
-   'body': '4',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 8,
-   'value': 4,
-   'entity_type': 'number'},
-  {'range': {'start': 0, 'end': 7},
-   'body': '2 hours',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 9,
-   'value': 7200,
-   'unit': 'hour',
-   'normalized': {'value': 7200, 'unit': 'second'},
-   '_meta': {},
-   'entity_type': 'duration'},
-  {'range': {'start': 10, 'end': 12},
-   'body': '$5',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 10,
-   'value': 5,
-   'unit': '$',
-   'entity_type': 'amount-of-money'},
-  {'range': {'start': 25, 'end': 41},
-   'body': '4111111111111111',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 11,
-   'entity_type': 'credit-card-number',
-   'issuer': 'visa',
-   'value': '4111111111111111'},
-  {'range': {'start': 25, 'end': 41},
-   'body': '4111111111111111',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.08333333333333333,
-   'alternative_index': 11,
-   'value': 4111111111111111,
-   'entity_type': 'number'}]}
-
    -
    -
    -
    -

    Testing

    -
      -

    1. Connect to the Duckling API either via a docker container or a local setup.

      2. -

    2. Boot an IPython session and setup an instance of DucklingPlugin.

      3. -
    -
     1from pprint import pprint
- 2from dialogy.workflow import Workflow
- 3from dialogy.base import Input
- 4from dialogy.plugins import DucklingPlugin
- 5
+
    This module exposes a parser for Duckling.
    
+
    Duckling helps parsing values like: date, time,
+numbers, currency etc. The parser will expect Duckling to be running as an http service, and provide
+means to connect from the implementation here.
    
+
     1from dialogy.workflow import Workflow
+ 2
+ 3def update_workflow(workflow, entities):
+ 4    workflow.input["duckling_entities"] = entities
+ 5
  6duckling_plugin = DucklingPlugin(
- 7    dest="output.entities",
- 8    dimensions=["people", "time"],
- 9    locale="en_IN",
-10    timezone="Asia/Kolkata",
-11    origin/master
-12)
-13
-14entities = duckling_plugin.parse(["We are 2 children coming tomorrow at 5 pm"])
-15pprint(entities)
-16# [PeopleEntity(
-17#     body='2 children', 
-18#     type='value', 
-19#     parsers=['DucklingPlugin'], 
-20#     score=1.0, 
-21#     alternative_index=0,
-22#     alternative_indices=[0, 0], 
-23#     latent=False, 
-24#     value=2, 
-25#     origin='value', 
-26#     unit='child',
-27#     entity_type='people')
-28# ,TimeEntity(
-29#     body='tomorrow at 5 pm', 
-30#     type='value', 
-31#     parsers=['DucklingPlugin'], 
-32#     score=1.0, 
-33#     alternative_index=0, 
-34#     alternative_indices=[0],
-35#     latent=False, 
-36#     value='2022-02-11T17:00:00.000+05:30', origin='value', entity_type='datetime', grain='hour')]
-37
-38# To convert these to dicts:
-39
-40entity_dicts = [entity.json() for entity in entities]
-41pprint(entity_dicts)
-42# [{'range': {'start': 7, 'end': 17},
-43#     'body': '2 children',
-44#     'type': 'value',
-45#     'parsers': ['DucklingPlugin'],
-46#     'score': 1.0,
-47#     'alternative_index': 0,
-48#     'value': 2,
-49#     'unit': 'child',
-50#     'entity_type': 'people'},
-51# {'range': {'start': 38, 'end': 54},
-52#     'body': 'tomorrow at 5 pm',
-53#     'type': 'value',
-54#     'parsers': ['DucklingPlugin'],
-55#     'score': 1.0,
-56#     'alternative_index': 0,
-57#     'value': '2022-02-11T17:00:00.000+05:30',
-58#     'entity_type': 'datetime',
-59#     'grain': 'hour'
-60# }]
-
    
-
    
-
    -
    -

    Guards

    -

    Like all plugins, the Duckling plugin can be guarded by conditions that prevent its execution. -You may need this if you want to save on latency and are sure of not expecting entities in those turns. We will use -the mother of all examples from above to make the point. We will use the current_state to tell the plugin -to guard itself in the “SMALL_TALK” state.

    -
    In [7]: from dialogy.workflow import Workflow
-   ...: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: 
-
-In [8]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=[
-   ...:         "number",
-   ...:         "people",
-   ...:         "date",
-   ...:         "time",
-   ...:         "duration",
-   ...:         "intersect",
-   ...:         "amount-of-money",
-   ...:         "credit-card-number",
-   ...:     ],
-   ...:     guards=[lambda i, o: i.current_state == "SMALL_TALK"]
-   ...: )
-   ...: 
-
-In [9]: workflow = Workflow([duckling_plugin])
-
-In [10]: utterances = [
-   ....:    "between today 7pm and tomorrow 9pm",
-   ....:    "can we get up at 6 am",
-   ....:    "we are 7 people",
-   ....:    "2 children 1 man 3 girls",
-   ....:    "can I come now?",
-   ....:    "can I come tomorrow",
-   ....:    "how about monday then",
-   ....:    "call me on 5th march",
-   ....:    "I want 4 pizzas",
-   ....:    "2 hours",
-   ....:    "can I pay $5 instead?",
-   ....:    "my credit card number is 4111111111111111",
-   ....: ]
-   ....: 
-
-In [11]: %%time
-   ....: input_, output = workflow.run(Input(utterances=utterances, current_state="SMALL_TALK"))
-   ....: 
-CPU times: user 92 us, sys: 23 us, total: 115 us
-Wall time: 117 us
-
-In [12]: output
-Out[12]: {'intents': [], 'entities': []}
-
    -
    -

    You can notice that the time drops from ms to us and we produce no entities.

    -
    -
    -

    Filtering Date and Time

    -

    There are cases where we run into ambiguity in entity resolution. A fun case happens usually with -the Hindi language when someone says “कल”. The problem is, “कल” could be “tommorrow” or “yesterday”.

    -
    -

    Problem

    -
    In [13]: import pytz
-   ....: from datetime import datetime
-   ....: from dialogy.workflow import Workflow
-   ....: from dialogy.base import Input
-   ....: from dialogy.plugins import DucklingPlugin
-   ....: from dialogy.utils import dt2timestamp
-   ....: 
-
-In [14]: duckling_plugin = DucklingPlugin(
-   ....:     dest="output.entities",
-   ....:     dimensions=["time"],
-   ....: )
-   ....: 
-
-In [15]: workflow = Workflow([duckling_plugin])
-
-In [16]: utterances = ["कल"]
-
-In [17]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ....: reference_time = dt2timestamp(reference_dt)
-   ....: 
-
-In [18]: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-In [19]: output
-Out[19]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 2},
-   'body': 'कल',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2021-12-31T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 0, 'end': 2},
-   'body': 'कल',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-02T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'}]}
-
    -
    -

    See the two entities?

    -
      -

    • We froze our reference time for 1st Jan 2022 at 12:00:00.

      • -

    • We get 31st December 2021

      • -

    • and 2nd Jan 2022.

      • -
    -

    This could be decoded using additional context. Let’s assume it is a given that dates can only be in future. Say -a flight-booking case.

    -
    In [20]: import pytz
-   ....: from datetime import datetime
-   ....: from dialogy.workflow import Workflow
-   ....: from dialogy.base import Input
-   ....: from dialogy.plugins import DucklingPlugin
-   ....: from dialogy.utils import dt2timestamp
-   ....: 
-
-In [21]: duckling_plugin = DucklingPlugin(
-   ....:     dest="output.entities",
-   ....:     dimensions=["time"],
-   ....:     datetime_filters=DucklingPlugin.FUTURE
-   ....: )
-   ....: 
-
-In [22]: workflow = Workflow([duckling_plugin])
-
-In [23]: utterances = ["कल"]
-
-In [24]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ....: reference_time = dt2timestamp(reference_dt)
-   ....: 
-
-In [25]: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-
-In [26]: output
-Out[26]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 2},
-   'body': 'कल',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-02T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'}]}
-
    -
    -

    We can use DucklingPlugin.PAST for inverse.

    -

    What if the context was within the utterance? Say for payment verification cases, we receive these utterances:

    -
      -

    • “कल ही कर दी थी” – “I have already paid yesterday”

      • -

    • “कल कर दूंगा” – “I will pay tomorrow”

      • -
    -

    We can’t filter out either future or past values. In these cases, the intent <Intents> carries the context -for resolving the entity value. We will look into Temporal Intents for more details.

    -
    -
    -
    -

    Temporal Intents

    -
    -

    Filtering

    -

    There are cases where intents carry a notion of time, repeating an example from above:

    -
      -

    1. “कल ही कर दी थी” – “I have already paid yesterday”

      2. -

    2. “कल कर दूंगा” – “I will pay tomorrow”

      3. -
    -

    While we can’t decode the entity value for “कल” in the above utterances, the correct intent prediction is:

    -
      -

    1. already_paid

      2. -

    2. pay_later

      3. -
    -

    If we could use this information, we could resolve the entity value for “कल”. We will simulate the output -to contain appropriate intents using the set method of workflow.

    -
    In [27]: import pytz
-   ....: from datetime import datetime
-   ....: from dialogy.workflow import Workflow
-   ....: from dialogy.base import Input
-   ....: from dialogy.types import Intent
-   ....: from dialogy.plugins import DucklingPlugin
-   ....: from dialogy.utils import dt2timestamp
-   ....: 
-
-In [28]: duckling_plugin = DucklingPlugin(
-   ....:     dest="output.entities",
-   ....:     dimensions=["time"],
-   ....:     temporal_intents={
-   ....:         "already_paid": DucklingPlugin.FUTURE,
-   ....:         "pay_later": DucklingPlugin.PAST
-   ....:     }
-   ....: )
-   ....: 
-
-In [29]: workflow = Workflow([duckling_plugin])
-
-In [30]: utterances = ["कल ही कर दी थी"]
-   ....: workflow.set("output.intents", [Intent(name="already_paid", score=1.0)])
-   ....: 
-Out[30]: Workflow(plugins=[<dialogy.plugins.text.duckling_plugin.DucklingPlugin object at 0x7fe13e70e940>], input=None, output=Output(intents=[Intent(name='already_paid', score=1.0, alternative_index=None)], entities=[]), debug=False)
-
-In [31]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ....: reference_time = dt2timestamp(reference_dt)
-   ....: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-   ....: 
-
-In [32]: output
-Out[32]: 
-{'intents': [{'name': 'already_paid',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': [{'range': {'start': 0, 'end': 2},
-   'body': 'कल',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-02T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'}]}
-
-In [33]: utterances = ["कल ही कर दी थी"]
-   ....: workflow.set("output.intents", [Intent(name="pay_later", score=1.0)])
-   ....: input_, output = workflow.run(Input(utterances=utterances, locale="hi_IN", reference_time=reference_time))
-   ....: 
-
-In [34]: output
-Out[34]: 
-{'intents': [{'name': 'pay_later',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': [{'range': {'start': 0, 'end': 2},
-   'body': 'कल',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2021-12-31T00:00:00.000+00:00',
-   'entity_type': 'date',
-   'grain': 'day'}]}
-
    -
    -
    -
    -

    Casting

    -

    We can also use temporal intents to cast certain entities. Machines can act on absolute time values but -natural language often has relative units like “for 2 hours”. These produce a duration entity.

    -
    In [35]: import pytz
-   ....: from datetime import datetime
-   ....: from dialogy.workflow import Workflow
-   ....: from dialogy.base import Input
-   ....: from dialogy.plugins import DucklingPlugin
-   ....: from dialogy.utils import dt2timestamp
-   ....: 
-
-In [36]: duckling_plugin = DucklingPlugin(
-   ....:     dest="output.entities",
-   ....:     dimensions=["duration", "time"],
-   ....: )
-   ....: 
-
-In [37]: workflow = Workflow([duckling_plugin])
-
-In [38]: utterances = ["for 2h"]
-
-In [39]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ....: reference_time = dt2timestamp(reference_dt)
-   ....: 
-
-In [40]: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-In [41]: output
-Out[41]: 
-{'intents': [],
- 'entities': [{'range': {'start': 4, 'end': 6},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': 7200,
-   'unit': 'hour',
-   'normalized': {'value': 7200, 'unit': 'second'},
-   '_meta': {},
-   'entity_type': 'duration'}]}
-
    -
    -

    We understand the duration but we could create an absolute time value using the reference time, but we don’t know if duration -should be added or removed.

    -
    In [42]: import pytz
-   ....: from datetime import datetime
-   ....: from dialogy.workflow import Workflow
-   ....: from dialogy.base import Input
-   ....: from dialogy.types import Intent
-   ....: from dialogy.plugins import DucklingPlugin
-   ....: from dialogy.utils import dt2timestamp
-   ....: 
-
-In [43]: duckling_plugin = DucklingPlugin(
-   ....:     dest="output.entities",
-   ....:     dimensions=["time", "duration"],
-   ....:     timezone="Asia/Kolkata",
-   ....:     temporal_intents={
-   ....:         "already_paid": DucklingPlugin.FUTURE,
-   ....:         "pay_later": DucklingPlugin.PAST
-   ....:     }
-   ....: )
-   ....: 
-
-In [44]: workflow = Workflow([duckling_plugin])
-
-In [45]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ....: reference_time = dt2timestamp(reference_dt)
-   ....: utterances = ["for 2h"]
-   ....: 
-
-In [46]: workflow.set("output.intents", [Intent(name="already_paid", score=1.0)])
-   ....: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-   ....: 
-
-In [47]: output
-Out[47]: 
-{'intents': [{'name': 'already_paid',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': [{'range': {'start': 4, 'end': 6},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-01T13:37:00+05:30',
-   'entity_type': 'datetime',
-   'grain': 'hour'}]}
-
-In [48]: workflow.set("output.intents", [Intent(name="pay_later", score=1.0)])
-   ....: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-   ....: 
-
-In [49]: output
-Out[49]: 
-{'intents': [{'name': 'pay_later',
-   'alternative_index': None,
-   'score': 1.0,
-   'parsers': [],
-   'slots': {}}],
- 'entities': [{'range': {'start': 4, 'end': 6},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-01T09:37:00+05:30',
-   'entity_type': 'datetime',
-   'grain': 'hour'}]}
-
    -
    -

    In case we need to always cast duration as a future or a past value we can do:

    -
    In [1]: import pytz
-   ...: from datetime import datetime
-   ...: from dialogy.workflow import Workflow
-   ...: from dialogy.base import Input
-   ...: from dialogy.types import Intent
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.utils import dt2timestamp
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time", "duration"],
-   ...:     timezone="Asia/Kolkata",
-   ...:     temporal_intents={
-   ...:         "__any__": DucklingPlugin.FUTURE,
-   ...:     }
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ...: reference_time = dt2timestamp(reference_dt)
-   ...: utterances = ["for 2h"]
-   ...: 
-
-In [5]: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-
-In [6]: output
-Out[6]: 
-{'intents': [],
- 'entities': [{'range': {'start': 4, 'end': 6},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-01T13:37:00+05:30',
-   'entity_type': 'datetime',
-   'grain': 'hour'}]}
-
-In [7]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time", "duration"],
-   ...:     timezone="Asia/Kolkata",
-   ...:     temporal_intents={
-   ...:         "__any__": DucklingPlugin.PAST,
-   ...:     }
-   ...: )
-   ...: 
-
-In [8]: workflow = Workflow([duckling_plugin])
-   ...: input_, output = workflow.run(Input(utterances=utterances, reference_time=reference_time))
-   ...: output
-   ...: 
-Out[8]: 
-{'intents': [],
- 'entities': [{'range': {'start': 4, 'end': 6},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': '2022-01-01T09:37:00+05:30',
-   'entity_type': 'datetime',
-   'grain': 'hour'}]}
+ 7    access=lambda workflow: (workflow.input["sentence"], workflow.input["reftime"], workflow.input["locale"])
+ 8    # the access method guides the plugin to data within a workflow.
+ 9    # in this case, the `input` property of a workflow is expected
+10    # to be a `dict` with a key named "sentence". Check line 30.
+11    mutate=update_workflow,
+12    # the mutate method guides the plugin to place the processed output
+13    # somewhere on the workflow.
+14    dimensions=["people"],
+15    # Duckling supports multiple dimensions, by specifying a list, we make sure
+16    # we are searching for a match within the expected dimensions only.
+17    locale="en_IN",
+18    # Duckling supports a set of locales, we need to provide this info to
+19    # get language specific matches.
+20    timezone="Asia/Kolkata",
+21    # Date/Time related entities make this field imperative.
+22)() # ⬅️ the instance is also getting called.
+23# Notice that we instantiate the class and also
+24# invoke __call__ method of the instance.
+25# The returned value from the instance is the plugin.
+26#
+27# This is a standard pattern across all plugins.
+28
+29workflow = Workflow(preprocessors=[duckling_plugin])
+30workflow.run(input_={"sentence": "there are 7 people"})
+31# Once an input is run through the workflow as on line 30,
+32# we can expect the plugin mutation to create an artifact
+33# as per line 4.
    -class DucklingPlugin(dimensions, timezone='UTC', timeout=0.5, url='http://0.0.0.0:8000/parse', locale='en_IN', temporal_intents=None, dest=None, guards=None, datetime_filters=None, threshold=None, activate_latent_entities=False, reference_time_column='reftime', input_column='alternatives', output_column=None, use_transform=False, debug=False)[source]
    +class DucklingPlugin(dimensions, timezone, timeout=0.5, url='http://0.0.0.0:8000/parse', locale='en_IN', datetime_filters=None, threshold=None, access=None, mutate=None, entity_map=None, activate_latent_entities=False, reference_time_column='reftime', input_column='alternatives', output_column=None, use_transform=False, debug=False)[source]

    Bases: dialogy.base.entity_extractor.EntityScoringMixin, dialogy.base.plugin.Plugin

    +

    A Plugin for extracting entities using Duckling. Once +instantiated, a duckling_parser object will interface to an http server, running Duckling.

    +

    This object when used as a plugin, transforms the List[Dict[str, Any]] returned from the API to a +BaseEntity or one of its subclasses.

    Parameters

    dimensions (List[str]) – Dimensions. Of the listed

    @@ -974,25 +314,6 @@

    Casting

    url (Optional[str]) – The address where Duckling’s entity parser can be reached, defaults to “http://0.0.0.0:8000/parse”.

    -
    -
    -FUTURE = 'future'
    -

    Select time entity values with scores greater than or equal to the reference time.

    -
    -
    -
    -PAST = 'past'
    -

    Select time entity values with scores lesser than or equal to the reference time.

    -
    -
    -
    -apply_entity_classes(list_of_entities, reference_time=None, timezone='UTC', duration_cast_operator=None)[source]
    -
    -
    Return type
    -

    List[BaseEntity]

    -
    -
    -
    apply_filters(entities)[source]
    @@ -1015,24 +336,6 @@

    Casting -
    -get_operator(filter_type)[source]
    -
    -
    Return type
    -

    Any

    -
    -
    -

    -
    -
    -parse(transcripts, locale=None, reference_time=None, use_latent=False, intents=None)[source]
    -
    -
    Return type
    -

    List[BaseEntity]

    -
    -
    -
    -
    select_datetime(entities, filter_type)[source]

    Select datetime entities as per the filters provided in the configuration.

    @@ -1069,11 +372,11 @@

    Casting
    -utility(input, output)[source]
    +utility(*args)[source]

    Produces Duckling entities, runs with a Workflow’s run method.

    Parameters
    -

    argbrightons – Expects a tuple of Tuple[natural language for parsing entities, reference time in milliseconds, locale]

    +

    args (Tuple(Union[str, List[str]], int, str)) – Expects a tuple of Tuple[natural language for parsing entities, reference time in milliseconds, locale]

    Returns

    A list of duckling entities.

    @@ -1083,18 +386,7 @@

    Casting -
    -validate(input_, reference_time)[source]
    -
    -
    Return type
    -

    DucklingPlugin

    -
    -

    -

    -
    -
    @@ -1150,21 +442,7 @@

    Casting diff --git a/docs/source/dialogy.plugins.text.html b/docs/source/dialogy.plugins.text.html index 810ec42a..4ef967f4 100644 --- a/docs/source/dialogy.plugins.text.html +++ b/docs/source/dialogy.plugins.text.html @@ -6,7 +6,7 @@ - dialogy.plugins.text package - dialogy 0.9.3 documentation + dialogy.plugins.text package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -236,48 +236,30 @@

    Subpackagesdialogy.plugins.text.classification package
  • dialogy.plugins.text.combine_date_time package
  • dialogy.plugins.text.duckling_plugin package
  • dialogy.plugins.text.lb_plugin package
  • dialogy.plugins.text.list_entity_plugin package
  • dialogy.plugins.text.list_search_plugin package
  • dialogy.plugins.text.merge_asr_output package

    • -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,37 +220,8 @@

    dialogy.plugins.text.lb_plugin package

    -
    -

    Module contents

    -
    -
    -class DucklingPluginLB(dimensions, timezone, timeout=0.5, url='http://0.0.0.0:8000/parse', locale='en_IN', datetime_filters=None, dest=None, guards=None, reference_time_column='reftime', input_column='alternatives', output_column=None, use_transform=False, debug=False)[source]
    -

    Bases: dialogy.plugins.text.duckling_plugin.DucklingPlugin

    -
    -
    -headers: Dict[str, str]
    -
    -
    -
    -reference_time: Optional[int]
    -
    -
    -
    -utility(input_, output)[source]
    -

    Produces Duckling entities, runs with a Workflow’s run method.

    -
    -
    Parameters
    -

    argbrightons – Expects a tuple of Tuple[natural language for parsing entities, reference time in milliseconds, locale]

    -
    -
    Returns
    -

    A list of duckling entities.

    -
    -
    Return type
    -

    List[BaseEntity]

    -
    -
    -
    -
    +
    +

    Module contents

    @@ -306,7 +277,7 @@

    dialogy.plugins.text.lb_plugin package diff --git a/docs/source/dialogy.plugins.text.list_entity_plugin.html b/docs/source/dialogy.plugins.text.list_entity_plugin.html index 8deeedef..c39781c2 100644 --- a/docs/source/dialogy.plugins.text.list_entity_plugin.html +++ b/docs/source/dialogy.plugins.text.list_entity_plugin.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.list_entity_plugin package - dialogy 0.9.3 documentation + dialogy.plugins.text.list_entity_plugin package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,90 +222,17 @@

    dialogy.plugins.text.list_entity_plugin package

    Module contents

    - -
    -

    Spacy NER

    -

    We also allow using spacy’s NER. This has to be passed within the -spacy_nlp attribute.

    -
    In [1]: import spacy
-   ...: from dialogy.workflow import Workflow
-   ...: from dialogy.plugins import ListEntityPlugin
-   ...: from dialogy.base import Input
-   ...: 
-
-In [2]: nlp = spacy.load("en_core_web_sm")
-
-In [3]: list_entity_plugin = ListEntityPlugin(
-   ...:   spacy_nlp=nlp,
-   ...:   style="spacy",
-   ...:   dest="output.entities"
-   ...: )
-   ...: workflow = Workflow([list_entity_plugin])
-   ...: _, output = workflow.run(Input(utterances="Need a place to stay in New Delhi."))
-   ...: 
-
-In [4]: output
-Out[4]: 
-{'intents': [],
- 'entities': [{'range': {'start': 24, 'end': 33},
-   'body': 'New Delhi',
-   'type': 'GPE',
-   'parsers': ['ListEntityPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': 'New Delhi',
-   'entity_type': 'GPE',
-   '_meta': {}}]}
-
    -
    +

    Module needs refactor. We are currently keeping all strategies bundled as methods as opposed to SearchStrategyClasses.

    +

    Within dialogy, we extract entities using Duckling, Pattern lists and Spacy. We can ship individual plugins but at the +same time, the difference is just configuration of each of these tools/services. There is another difference of +intermediate structure that the DucklingPlugin expects. We need to prevent the impact of the structure from affecting +all other entities. So that their from_dict(...) methods are pristine and involve no shape hacking.

    -class ListEntityPlugin(style=None, candidates=None, spacy_nlp=None, dest=None, guards=None, labels=None, threshold=None, input_column='alternatives', output_column=None, use_transform=True, flags=RegexFlag.None, debug=False)[source]
    +class ListEntityPlugin(style=None, candidates=None, spacy_nlp=None, labels=None, threshold=None, access=None, mutate=None, input_column='alternatives', output_column=None, use_transform=True, flags=RegexFlag.None, debug=False)[source]

    Bases: dialogy.base.entity_extractor.EntityScoringMixin, dialogy.base.plugin.Plugin

    -

    A Plugin for extracting entities using spacy or a list of regex patterns.

    +

    A Plugin for extracting entities using spacy or a list of regex patterns.

    Parameters
    @@ -313,10 +240,14 @@

    Spacy NERdict that shows a mapping of entity values and their patterns.

    -

  • spacy_nlp (Any) – Required if style is “spacy”, requires is a +

  • spacy_nlp (Any) – Required if style is “spacy”, this is a spacy model.

  • labels (Optional[List[str]]) – Required if style is “spacy”. If there is a need to extract only a few labels from all the other available labels.

    • +

  • access (Optional[PluginFn]) – A plugin io utility that allows access to transcripts +List[str] within a Workflow.

    • +

  • mutate (Optional[PluginFn]) – A plugin io utility that allows insertion of List[BaseEntity] within a +Workflow.

  • debug (bool) – A flag to set debugging on the plugin methods

    • @@ -333,7 +264,7 @@

    Spacy NER

    List of entities from regex matches or spacy ner.

    Return type
    -

    List[KeywordEntity]

    +

    List[KeywordEntity]

    @@ -387,26 +318,20 @@

    Spacy NER
    -utility(input, _)[source]
    -

    An abstract method that describes the plugin’s functionality.

    +utility(*args)[source] +

    Transform X -> y.

    +

    This method is called during a Workflow run.

    -
    Parameters
    -
      -

    • input (Input) – The workflow’s input.

      • -

    • output (Output) – The workflow’s output.

      • -
    +
    Returns
    +

    A value that will be passed to the next plugin (if any).

    -
    Returns
    -

    The value returned by the plugin.

    -
    -
    Return type
    -

    Any

    +
    Return type
    +

    Any

    -
    @@ -461,11 +386,7 @@

    Spacy NER diff --git a/docs/source/dialogy.plugins.text.list_search_plugin.html b/docs/source/dialogy.plugins.text.list_search_plugin.html index 4975e02d..6e9399a9 100644 --- a/docs/source/dialogy.plugins.text.list_search_plugin.html +++ b/docs/source/dialogy.plugins.text.list_search_plugin.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.list_search_plugin package - dialogy 0.9.3 documentation + dialogy.plugins.text.list_search_plugin package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,107 +220,8 @@

    dialogy.plugins.text.list_search_plugin package

    -
    -

    Module contents

    -

    Module needs refactor. We are currently keeping all strategies bundled as methods as opposed to SearchStrategyClasses.

    -

    Within dialogy, we extract entities using Duckling, Pattern lists and Spacy. We can ship individual plugins but at the -same time, the difference is just configuration of each of these tools/services. There is another difference of -intermediate structure that the DucklingPlugin expects. We need to prevent the impact of the structure from affecting -all other entities. So that their from_dict(...) methods are pristine and involve no shape hacking.

    -
    -
    -class ListSearchPlugin(fuzzy_dp_config, threshold=None, dest=None, guards=None, input_column='alternatives', output_column=None, use_transform=True, flags=RegexFlag.None, debug=False, fuzzy_threshold=0.1)[source]
    -

    Bases: dialogy.base.entity_extractor.EntityScoringMixin, dialogy.base.plugin.Plugin

    -
    -

    A Plugin for extracting entities using spacy or a list of regex patterns.

    -
    -
    -
    Parameters
    -
      -

    • style (Optional[str]) – One of [“regex”, “spacy”]

      • -

    • candidates (Optional[Dict[str, List[str]]]) – Required if style is “regex”, this is a dict that shows a mapping of entity -values and their patterns.

      • -

    • spacy_nlp (Any) – Required if style is “spacy”, this is a -spacy model.

      • -

    • labels (Optional[List[str]]) – Required if style is “spacy”. If there is a need to extract only a few labels from all the other -available labels.

      • -

    • debug (bool) – A flag to set debugging on the plugin methods

      • -
    -
    -
    -
    - -
    -
    Return type
    -

    Tuple[str, str, str, Tuple[int, int], float]

    -
    -
    -
    -
    -
    -fuzzy_init()[source]
    -

    Initializing the parameters for fuzzy dp search with their values

    -
    -
    Return type
    -

    None

    -
    -
    -
    -
    -
    -get_entities(transcripts, lang)[source]
    -

    Parse entities using regex and spacy ner.

    -
    -
    Parameters
    -

    transcripts (List[str]) – A list of strings within which to search for entities.

    -
    -
    Returns
    -

    List of entities from regex matches or spacy ner.

    -
    -
    Return type
    -

    List[KeywordEntity]

    -
    -
    -
    -
    - -

    Search for Entity in transcript from a defined List Search space -:param transcripts : A list of transcripts, List[str]. -:param lang : Language code of the transcript :code str -:return: Token matches with the transcript. -:rtype: List[MatchType]

    -
    -
    -
    -search_regex(query, entity_type='', entity_patterns=[''], match_dict={})[source]
    -
    -
    Return type
    -

    Tuple[str, str, str, Tuple[int, int], float]

    -
    -
    -
    -
    -
    -utility(input_, _)[source]
    -

    An abstract method that describes the plugin’s functionality.

    -
    -
    Parameters
    -
      -

    • input (Input) – The workflow’s input.

      • -

    • output (Output) – The workflow’s output.

      • -
    -
    -
    Returns
    -

    The value returned by the plugin.

    -
    -
    Return type
    -

    Any

    -
    -
    -
    -
    +
    +

    Module contents

    @@ -376,7 +277,7 @@

    dialogy.plugins.text.list_search_plugin package diff --git a/docs/source/dialogy.plugins.text.merge_asr_output.html b/docs/source/dialogy.plugins.text.merge_asr_output.html index d5f8eb4b..047daad9 100644 --- a/docs/source/dialogy.plugins.text.merge_asr_output.html +++ b/docs/source/dialogy.plugins.text.merge_asr_output.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.merge_asr_output package - dialogy 0.9.3 documentation + dialogy.plugins.text.merge_asr_output package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,40 +222,25 @@

    dialogy.plugins.text.merge_asr_output package

    Module contents

    -

    We use classifiers for prediction of intents, utterances from an ASR are prominent features -for the task. We featurize the utterances by concatenation of all transcripts.

    -
    In [1]: from dialogy.workflow import Workflow
-   ...: from dialogy.plugins import MergeASROutputPlugin
-   ...: from dialogy.base import Input
-   ...: 
-
-In [2]: merge_asr_output_plugin = MergeASROutputPlugin(dest="input.clf_feature")
-   ...: workflow = Workflow([merge_asr_output_plugin])
-   ...: 
-
-In [3]: input_, _ = workflow.run(Input(utterances=["we will come by 7 pm", " will come by 7 pm"]))
-
-In [4]: input_
-Out[4]: 
-{'utterances': ['we will come by 7 pm', ' will come by 7 pm'],
- 'reference_time': None,
- 'latent_entities': False,
- 'transcripts': ['we will come by 7 pm', ' will come by 7 pm'],
- 'clf_feature': ['<s> we will come by 7 pm </s> <s>  will come by 7 pm </s>'],
- 'lang': 'en',
- 'locale': 'en_IN',
- 'timezone': 'UTC',
- 'slot_tracker': None,
- 'current_state': None,
- 'previous_intent': None,
- 'history': None}
-
    -
    +

    We expect ASR transcripts to be the most common inputs to the system. Since there is a chance that we can improve the +performance by asking the ASR to provide more information in the form of multiple alternatives. The problem though, is +they can’t be used as it is. Models that can operate on text data, need them to be vectorized. These vectors have +semantic significance, where different text inputs imply different vectors. If their semantic differences are modeled +well, we can expect these vectors to be similar or dissimilar just like the text that was used to create them.

    +

    The problem with transcripts is, there will be noise in the transcripts, since we asked for N transcripts, +by definition, an ASR will provide N unique transcripts. The noise may in fact be inconsequential for downstream NLP +tasks, but we have noticed that there are certain inputs owing to audio artifacts or state of LMs tend to have a great +impact on the transcription.

    +

    This module will ship a simple concatenation strategy to address this problem. The NLP model should be expected to +handle documents (List[str]) instead of single text. The plugin then will be able to produce a str by +simply concatenation of each transcript.

    +

    This may not be very helpful, there is a VotePlugin under development so use it with caution. It may +help with precision at the cost of recall.

    -class MergeASROutputPlugin(input_column='alternatives', output_column=None, use_transform=False, dest=None, guards=None, debug=False)[source]
    +class MergeASROutputPlugin(access, mutate, input_column='alternatives', output_column=None, use_transform=False, debug=False)[source]

    Bases: dialogy.base.plugin.Plugin

    -

    Working details are covered in MergeASROutputPlugin.

    +

    Working details are covered in merge_asr_output.

    Parameters

    Plugin ([type]) – [description]

    @@ -273,20 +258,15 @@

    dialogy.plugins.text.merge_asr_output package
    -utility(input, _)[source]
    -

    An abstract method that describes the plugin’s functionality.

    +utility(*args)[source] +

    Transform X -> y.

    +

    This method is called during a Workflow run.

    -
    Parameters
    -
      -

    • input (Input) – The workflow’s input.

      • -

    • output (Output) – The workflow’s output.

      • -
    +
    Returns
    +

    A value that will be passed to the next plugin (if any).

    -
    Returns
    -

    The value returned by the plugin.

    -
    -
    Return type
    -

    Any

    +
    Return type
    +

    Any

    @@ -303,7 +283,15 @@

    dialogy.plugins.text.merge_asr_output packagenormalize

    +

    The normalization is done by normalize

    +
    In [1]: from dialogy.plugins.text.merge_asr_output import merge_asr_output
+
+In [2]: utterances = ["This is a sentence", "That is a sentence"]
+
+In [3]: merge_asr_output(utterances)
+Out[3]: ['<s> this is a sentence </s> <s> that is a sentence </s>']
+
    +
    Parameters

    utterances (Any) –

    A structure representing ASR output. We support only:

    diff --git a/docs/source/dialogy.plugins.text.slot_filler.html b/docs/source/dialogy.plugins.text.slot_filler.html index efa553e2..1e46e558 100644 --- a/docs/source/dialogy.plugins.text.slot_filler.html +++ b/docs/source/dialogy.plugins.text.slot_filler.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.slot_filler package - dialogy 0.9.3 documentation + dialogy.plugins.text.slot_filler package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -225,27 +225,96 @@

    Submodules

    dialogy.plugins.text.slot_filler.rule_slot_filler module

    -
    -

    Slot Filling

    -

    While we describe slots in the linked doc. The slot filling is part of an intent’s -role. This plugin orchestrates the execution of slot filling via methods available on intents.

    -
      -

    1. We eagerly apply slot placeholders using the apply method.

      2. -

    2. Once entities are found, we check if the slots can fill them using the fill method.

      3. -

    3. If no slots were filled, we remove the placeholders using cleanup method.

      4. -
    +

    Module provides access to a rule-based slot filler.

    -class RuleBasedSlotFillerPlugin(rules, dest=None, guards=None, replace_output=True, fill_multiple=True, debug=False)[source]
    +class RuleBasedSlotFillerPlugin(rules, fill_multiple=False, access=None, mutate=None, debug=False)[source]

    Bases: dialogy.base.plugin.Plugin

    -

    A utility plugin for +

    A utility plugin for slot filling <https://nlpprogress.com/english/intent_detection_slot_filling.html>._ An Intent may have a few slots that need to be filled.

    This plugin can assist filling pertaining to certain intent:entity:slot-name association rules.

    -

    Schema for rules

    +

    Schema for rules looks like:

    +
    {
+    "intent_name": {
+        "slot_name":"entity_type"
+    }
+}
+
    +
    +

    This can be represented in a yaml format like so:

    +
    slots:
+    faqs:
+        action_slot: actions
+    report:
+        time_slot:
+        - time
+        - number
+
    +

    Let’s run through a practical example. We will create a workflow and preset the output to have expected intent and entities.

    -
    +
    In [1]: import yaml
+
+In [2]: from pprint import pprint
+
+In [3]: from dialogy.workflow import Workflow
+
+In [4]: from dialogy.plugins import RuleBasedSlotFillerPlugin
+
+In [5]: from dialogy.types.intent import Intent
+
+In [6]: from dialogy.types.entity import PeopleEntity, NumericalEntity
+
+# ---------------------------------------------------------------------------------------------------
+# Create rules
+In [7]: rule_string = """
+   ...: faqs:
+   ...:     number_slot: number
+   ...: count_people:
+   ...:     number_slot:
+   ...:     - people
+   ...:     - number
+   ...: """
+   ...: 
+
+In [8]: slot_rules = yaml.safe_load(rule_string)
+
+# ---------------------------------------------------------------------------------------------------
+# ---------------------------------------------------------------------------------------------------
+# Setting up the plugin
+In [9]: slot_filler = RuleBasedSlotFillerPlugin(rules=slot_rules, access=lambda w: w.output, debug=True)
+
+# ---------------------------------------------------------------------------------------------------
+# ---------------------------------------------------------------------------------------------------
+# Creating synthetic values.
+# normally you'd get these by running the workflow with models.
+In [10]: intent = Intent(name="count_people", score=1)
+
+In [11]: number_entity = NumericalEntity(type="number", range={"start": 0, "end": 7}, body='2 people', latent=False, values=[{"values":2}], dim="number")
+
+In [12]: people_entity = PeopleEntity(type="people", range={"start": 0, "end": 7}, body='2 people', latent=False, values=[{"values":2}], dim="people")
+
+In [13]: entities = [people_entity, number_entity]
+
+# ---------------------------------------------------------------------------------------------------
+# ---------------------------------------------------------------------------------------------------
+# setting up the workflow to use slot_filler
+In [14]: workflow = Workflow([slot_filler], debug=True)
+
+# ---------------------------------------------------------------------------------------------------
+# If you notice the instantiation of RuleBasedSlotFillerPlugin,
+# you will notice the `access` method expects a Tuple of two elements.
+In [15]: workflow.output = (intent, entities)
+
+In [16]: workflow.run(input_="2 am")
+Out[16]: 
+(Intent(name=count_people, score=1, slots={}),
+ [PeopleEntity(type='people', body='2 people', parsers=[], score=None, alternative_index=None, alternative_indices=None, latent=False, value=None, origin='value', unit=''),
+  NumericalEntity(type='number', body='2 people', parsers=[], score=None, alternative_index=None, alternative_indices=None, latent=False, value=None, origin='value')])
+
    +
    +
    Parameters

    • rules (Rule) – A mapping that defines relationship between an intent, its slots and the entities that fill them.

      • @@ -261,32 +330,26 @@

      Slot Fillingfill(intents, entities)[source]
      Return type
      -

      List[Intent]

      +

      None

    -utility(_, output)[source]
    -

    An abstract method that describes the plugin’s functionality.

    +utility(*args)[source] +

    Transform X -> y.

    +

    This method is called during a Workflow run.

    -
    Parameters
    -
      -

    • input (Input) – The workflow’s input.

      • -

    • output (Output) – The workflow’s output.

      • -
    +
    Returns
    +

    A value that will be passed to the next plugin (if any).

    -
    Returns
    -

    The value returned by the plugin.

    -
    -
    Return type
    -

    Any

    +
    Return type
    +

    Any

    -

    Module contents

    To handle a single dialog we need a system that can:

    @@ -296,12 +359,12 @@

    Slot FillingIntent and a set of -Entity are identified, there needs to be an arrangement that +Entity are identified, there needs to be an arrangement that tells us if any of these are related to each other.

    As per the current design, an Intent can contain a number of Slot. A Slot fills an entity of a pre-defined type.

      -

    • A Slot configured to fill a TimeEntity will not fill a KeywordEntity.

      • +

    • A Slot configured to fill a TimeEntity will not fill a LocationEntity.

    • A Slot will not fill more than one entity even if the type matches, unless fill_multiple is provided.

    @@ -360,10 +423,7 @@

    Slot Fillingdialogy.plugins.text.slot_filler package diff --git a/docs/source/dialogy.plugins.text.voting.html b/docs/source/dialogy.plugins.text.voting.html index 67daa5fe..5c97418f 100644 --- a/docs/source/dialogy.plugins.text.voting.html +++ b/docs/source/dialogy.plugins.text.voting.html @@ -6,7 +6,7 @@ - dialogy.plugins.text.voting package - dialogy 0.9.3 documentation + dialogy.plugins.text.voting package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation diff --git a/docs/source/dialogy.types.entity.amount_of_money.html b/docs/source/dialogy.types.entity.amount_of_money.html index a91720e6..1efbe523 100644 --- a/docs/source/dialogy.types.entity.amount_of_money.html +++ b/docs/source/dialogy.types.entity.amount_of_money.html @@ -6,7 +6,7 @@ - dialogy.types.entity.amount_of_money package - dialogy 0.9.3 documentation + dialogy.types.entity.amount_of_money package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,67 +220,8 @@

    dialogy.types.entity.amount_of_money package

    -
    -

    Module contents

    -

    Module provides access to entity types that can be parsed to currencies and their value.

    -
    -
    Import classes:
      -

    • CurrencyEntity

      • -
    -
    -
    -
    -
    -class CurrencyEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, *, range, type='value', unit, entity_type='amount-of-money')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    Numerical Entity Type

    -

    Use this type for handling all entities that can be parsed to obtain: -- numbers -- date -- time -- datetime

    -
    -
    Attributes:
      -

    • dim dimension of the entity from duckling parser

      • -

    • type is the type of the entity which can have values in [“value”, “interval”]

      • -
    -
    -
    -
    -
    -entity_type: str
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    CurrencyEntity

    -
    -
    -
    -
    -
    -get_value()[source]
    -

    Getter for CurrencyEntity.

    -

    We are yet to decide the pros and cons of the output. It seems retaining {“value”: float, “unit”: }

    -
    -
    Parameters
    -

    reference (Any, optional) – [description], defaults to None

    -
    -
    Returns
    -

    [description]

    -
    -
    Return type
    -

    Any

    -
    -
    -
    -
    -
    -unit: str
    -
    -
    +
    +

    Module contents

    @@ -336,7 +277,7 @@

    dialogy.types.entity.amount_of_money package diff --git a/docs/source/dialogy.types.entity.base_entity.html b/docs/source/dialogy.types.entity.base_entity.html index 29b3e782..a698fee1 100644 --- a/docs/source/dialogy.types.entity.base_entity.html +++ b/docs/source/dialogy.types.entity.base_entity.html @@ -6,7 +6,7 @@ - dialogy.types.entity.base_entity package - dialogy 0.9.3 documentation + dialogy.types.entity.base_entity package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,78 +222,17 @@

    dialogy.types.entity.base_entity package

    Module contents

    -

    Module provides access to a base class to create other entities.

    -
    - classDiagram - direction LR - BaseEntity --|> NumericalEntity - BaseEntity --|> CreditCardNumberEntity - BaseEntity --|> CurrencyEntity - BaseEntity --|> TimeEntity - BaseEntity --|> KeywordEntity - BaseEntity --|> DurationEntity - BaseEntity --|> PeopleEntity - - TimeEntity --|> TimeIntervalEntity - - NumericalEntity ..> TimeEntity: as_time - DurationEntity ..> TimeEntity: as_time - - class BaseEntity { - +Dict[str, int] range - +str body - +str dim - +int alternative_index - +str type - +str entity_type - +float score - +Any value - +List~Any~ values - +List~str~ parsers - +bool latent - +Any get_value() - +Dict[str, Any] json() - $BaseEntity from_dict(d) - } - - class CurrencyEntity { - +str unit - +Any get_value() - +CurrencyEntity from_duckling(d) - } - - class CreditCardNumberEntity { - +str issuer - +from_duckling(d: dict, idx: int) - +CreditCardNumberEntity from_duckling(d) - } - - class KeywordEntity { - } - - class DurationEntity { - +str unit - +dict normalized - +DurationEntity from_duckling(d) - } - - class NumericalEntity { - +NumericalEntity from_duckling(d) - } - - class PeopleEntity { - +str unit - +PeopleEntity from_duckling(d) - } - - class TimeEntity { - +str grain - +datetime get_value() - +TimeEntity from_duckling(d) - } -
    +

    Module provides access to a base class to create other entities.

    +

    These methods of this class that are supposed to be overridden: +- validate +- get_value

    +

    Import classes:

    +
      +

    • BaseEntity

      • +
    +
    -class BaseEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, entity_type=None, *, range, type='value')[source]
    +class BaseEntity(range, type, body, dim=None, parsers=NOTHING, score=None, slot_names=NOTHING, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, entity_type=None)[source]

    Bases: object

    Base Entity Type.

    This class is meant for subclassing. @@ -316,18 +255,6 @@

    dialogy.types.entity.base_entity package -
    -alternative_index: Optional[int]
    -

    -
    -
    -alternative_indices: Optional[List[int]]
    -
    -
    -
    -body: str
    -
    copy()[source]
    @@ -337,25 +264,31 @@

    dialogy.types.entity.base_entity package -
    -dim: Optional[str]
    -

    -
    entity_type: Optional[str]
    -classmethod from_dict(dict_, reference=None)[source]
    -
    +classmethod from_dict(dict_)[source] +

    Create an instance of a given class cls from a dict that complies +with attributes of cls through its keys and values. +Compliance is verified using the __validate method. It is expected that each subclass +will implement their own flavor of __validate to check their respective inputs.

    +
    +
    Parameters
    +

    dict (Dict[str, Any]) – A dict that provides all the attributes necessary for instantiating this class

    +
    +
    Returns
    +

    Instance of class

    +
    Return type
    -

    BaseEntity

    +

    BaseEntity

    -get_value()[source]
    +get_value(reference=None)[source]

    Get value of an entity.

    The structure of an entity conceals the value in different data structures. This is sugar for access.

    @@ -400,34 +333,51 @@

    dialogy.types.entity.base_entity package -
    -latent: bool
    -

    -
    -
    -parsers: List[str]
    -
    -
    -
    -range: Dict[str, int]
    -
    -
    -
    -score: Optional[float]
    -
    -
    -
    -type: str
    -
    +
    +
    +classmethod reshape(dict_)[source]
    +
    +
    Return type
    +

    Dict[str, Any]

    +
    +
    +
    +
    +
    +set_value(value=None)[source]
    +

    Set values and value attribute.

    +
    +
    Parameters
    +

    value (Any) – The parsed value of an entity token.

    +
    +
    Returns
    +

    None

    +
    +
    Return type
    +

    None

    +
    +
    +
    +
    +
    +classmethod validate(dict_)[source]
    +

    Check attributes of instance match expected types.

    +
    +
    Parameters
    +

    dict (Dict[str, Any]) – A Dict where each key is an attribute of the instance and value is the expected type.

    +
    +
    Returns
    +

    None

    +
    +
    Return type
    +

    None

    +
    +
    +
    value: Any
    -
    -
    -values: List[Dict[str, Any]]
    -
    diff --git a/docs/source/dialogy.types.entity.credit_card_number.html b/docs/source/dialogy.types.entity.credit_card_number.html index 64851389..5ccc86ca 100644 --- a/docs/source/dialogy.types.entity.credit_card_number.html +++ b/docs/source/dialogy.types.entity.credit_card_number.html @@ -6,7 +6,7 @@ - dialogy.types.entity.credit_card_number package - dialogy 0.9.3 documentation + dialogy.types.entity.credit_card_number package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,93 +220,8 @@

    dialogy.types.entity.credit_card_number package

    -
    -

    Module contents

    -
    -

    Credit Card Number Entity

    -

    Provides the entity class for credit card numbers in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["credit-card-number"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("my credit card number is 4111111111111111.")
-Out[3]: [CreditCardNumberEntity(body='4111111111111111', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, issuer='visa', value='4111111111111111', values=[{'value': '4111111111111111'}])]
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["credit-card-number"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: _, output = workflow.run(Input(utterances="my credit card number is 4111111111111111."))
-
-In [5]: output
-Out[5]: 
-{'intents': [],
- 'entities': [{'range': {'start': 25, 'end': 41},
-   'body': '4111111111111111',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'entity_type': 'credit-card-number',
-   'issuer': 'visa',
-   'value': '4111111111111111'}]}
-
    -
    -
    -
    -class CreditCardNumberEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, *, range, type='value', entity_type='credit-card-number', issuer, value=None, values)[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -
    -
    -entity_type: Optional[str]
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    CreditCardNumberEntity

    -
    -
    -
    -
    -
    -issuer: str
    -
    -
    -
    -value: Optional[str]
    -
    -
    -
    -values: List[Dict[str, Any]]
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -362,14 +277,7 @@

    Workflow Integration diff --git a/docs/source/dialogy.types.entity.deserialize.html b/docs/source/dialogy.types.entity.deserialize.html index b40cc199..2fb57cad 100644 --- a/docs/source/dialogy.types.entity.deserialize.html +++ b/docs/source/dialogy.types.entity.deserialize.html @@ -6,7 +6,7 @@ - dialogy.types.entity.deserialize package - dialogy 0.9.3 documentation + dialogy.types.entity.deserialize package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,17 +220,8 @@

    dialogy.types.entity.deserialize package

    -
    -

    Module contents

    -
    -
    -deserialize_duckling_entity(duckling_entity_dict, alternative_index, reference_time=None, timezone='UTC', duration_cast_operator=None)[source]
    -
    -
    Return type
    -

    BaseEntity

    -
    -
    -
    +
    +

    Module contents

    @@ -268,7 +259,7 @@

    dialogy.types.entity.deserialize package diff --git a/docs/source/dialogy.types.entity.duration.html b/docs/source/dialogy.types.entity.duration.html index 6b16f1b6..4910eba9 100644 --- a/docs/source/dialogy.types.entity.duration.html +++ b/docs/source/dialogy.types.entity.duration.html @@ -6,7 +6,7 @@ - dialogy.types.entity.duration package - dialogy 0.9.3 documentation + dialogy.types.entity.duration package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,107 +220,8 @@

    dialogy.types.entity.duration package

    -
    -

    Module contents

    -
    -

    Duration Entity

    -

    Provides the entity class for durations in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["duration"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("she said 2h.")
-Out[3]: [DurationEntity(body='2h', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value=7200, unit='hour', normalized={'value': 7200, 'unit': 'second'}, _meta={}, entity_type='duration')]
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["duration"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: _, output = workflow.run(Input(utterances="she said 2h"))
-
-In [5]: output
-Out[5]: 
-{'intents': [],
- 'entities': [{'range': {'start': 9, 'end': 11},
-   'body': '2h',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 1.0,
-   'alternative_index': 0,
-   'value': 7200,
-   'unit': 'hour',
-   'normalized': {'value': 7200, 'unit': 'second'},
-   '_meta': {},
-   'entity_type': 'duration'}]}
-
    -
    -
    -
    -class DurationEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, normalized=NOTHING, meta=NOTHING, *, range, type='value', unit, entity_type='duration')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    This entity type expects a normalized attribute. This provides the duration normalized to seconds.

    -

    Helpful in cases where we wish to operate on time like: -“I want a booking in 2 days.”

    -

    We can tell the time at which the sentence was said, but we need to make the booking after two days.

    -

    This entity parses this information and also provides us the number of seconds to add to the current timestamp -to get to a date that’s 2 days ahead.

    -
    -
    -as_time(reference_unix_ts, timezone, duration_cast_operator)[source]
    -

    Converts a duration entity to a time entity.

    -
    -
    Return type
    -

    TimeEntity

    -
    -
    -
    -
    -
    -entity_type: str
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    DurationEntity

    -
    -
    -
    -
    -
    -normalized: Dict[str, Any]
    -
    -
    -
    -unit: str
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -376,14 +277,7 @@

    Workflow Integration diff --git a/docs/source/dialogy.types.entity.html b/docs/source/dialogy.types.entity.html index cc3feb39..39128cb6 100644 --- a/docs/source/dialogy.types.entity.html +++ b/docs/source/dialogy.types.entity.html @@ -6,7 +6,7 @@ - dialogy.types.entity package - dialogy 0.9.3 documentation + dialogy.types.entity package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -225,7 +225,7 @@

    Subpackages

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,71 +220,8 @@

    dialogy.types.entity.keyword package

    -
    -

    Module contents

    -

    Module provides access to an entity type (class) to contain keyword extraction based entites. -These entities originate from token lookups, regex, etc.

    -
    -
    Import classes:
      -

    • KeywordEntity

      • -
    -
    -
    -
    -
    -class KeywordEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, entity_type=None, meta=NOTHING, *, range, type='value')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    Use this type for handling keyword based extractions where presence of specific tokens in the ASR -is enough for detection.

    -
    -
    -alternative_index: Optional[int]
    -
    -
    -
    -alternative_indices: Optional[List[int]]
    -
    -
    -
    -body: str
    -
    -
    -
    -dim: Optional[str]
    -
    -
    -
    -entity_type: Optional[str]
    -
    -
    -
    -latent: bool
    -
    -
    -
    -parsers: List[str]
    -
    -
    -
    -range: Dict[str, int]
    -
    -
    -
    -score: Optional[float]
    -
    -
    -
    -type: str
    -
    -
    -
    -value: Any
    -
    -
    -
    -values: List[Dict[str, Any]]
    -
    -
    +
    +

    Module contents

    @@ -340,7 +277,7 @@

    dialogy.types.entity.keyword package diff --git a/docs/source/dialogy.types.entity.numerical.html b/docs/source/dialogy.types.entity.numerical.html index 73c2dcc4..7501417b 100644 --- a/docs/source/dialogy.types.entity.numerical.html +++ b/docs/source/dialogy.types.entity.numerical.html @@ -6,7 +6,7 @@ - dialogy.types.entity.numerical package - dialogy 0.9.3 documentation + dialogy.types.entity.numerical package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,146 +220,8 @@

    dialogy.types.entity.numerical package

    -
    -

    Module contents

    -
    -

    Numerical Entity

    -

    Provides the entity class for representing numbers in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["number"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("12")
-Out[3]: [NumericalEntity(body='12', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value=12, origin='value', entity_type='number')]
-
-In [4]: duckling_plugin.parse("twenty two")
-Out[4]: [NumericalEntity(body='twenty two', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value=22, origin='value', entity_type='number')]
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [5]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [6]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["number"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [7]: workflow = Workflow([duckling_plugin])
-
-In [8]: _, output = workflow.run(Input(utterances=["seventy two", "4 fruits"]))
-
-In [9]: output
-Out[9]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 11},
-   'body': 'seventy two',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.5,
-   'alternative_index': 0,
-   'value': 72,
-   'entity_type': 'number'},
-  {'range': {'start': 0, 'end': 1},
-   'body': '4',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.5,
-   'alternative_index': 1,
-   'value': 4,
-   'entity_type': 'number'}]}
-
    -
    -
    -
    -

    CASTING

    -

    We have noticed that there is a need to cast numerical values to time. The as_time method can be used to do this. -The default assumption is to use the number as day but we can also replace the month and hour.

    -
    In [1]: import pytz
-   ...: from datetime import datetime
-   ...: from dialogy.workflow import Workflow
-   ...: from dialogy.base import Input
-   ...: from dialogy.types import Intent
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.utils import dt2timestamp
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["number"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: reference_dt = datetime(2022, 1, 1, 12, 0, 0, tzinfo=pytz.timezone("Asia/Kolkata"))
-   ...: reference_time = dt2timestamp(reference_dt)
-   ...: 
-
-In [4]: entities = duckling_plugin.parse("12")
-
-In [5]: [entity.as_time(reference_time, "Asia/Kolkata") for entity in entities]
-Out[5]: [TimeEntity(body='12', type='value', parsers=[], score=None, alternative_index=0, alternative_indices=None, latent=False, value='2022-01-12T11:37:00+05:30', entity_type='date', origin='value', dim='time', grain='day')]
-
    -
    -
    -
    -class NumericalEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, origin='value', entity_type='number', *, range, type='value')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    Numerical Entity Type

    -

    Use this type for handling all entities that can be parsed to obtain: -- numbers -- date -- time -- datetime

    -
    -
    Attributes:
      -

    • dim dimension of the entity from duckling parser

      • -

    • type is the type of the entity which can have values in [“value”, “interval”]

      • -
    -
    -
    -
    -
    -as_time(reference_unix_ts, timezone, replace='day')[source]
    -

    Converts a duration entity to a time entity.

    -
    -
    Return type
    -

    TimeEntity

    -
    -
    -
    -
    -
    -entity_type: Optional[str]
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    NumericalEntity

    -
    -
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -415,15 +277,7 @@

    CASTING diff --git a/docs/source/dialogy.types.entity.people.html b/docs/source/dialogy.types.entity.people.html index b562a5e2..8e539347 100644 --- a/docs/source/dialogy.types.entity.people.html +++ b/docs/source/dialogy.types.entity.people.html @@ -6,7 +6,7 @@ - dialogy.types.entity.people package - dialogy 0.9.3 documentation + dialogy.types.entity.people package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,76 +220,8 @@

    dialogy.types.entity.people package

    -
    -

    Module contents

    -
    -

    People Entity

    -

    Provides the entity class for representing number of people in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["people"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("two kids")
-Out[3]: []
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["people"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: _, output = workflow.run(Input(utterances="two kids"))
-
-In [5]: output
-Out[5]: {'intents': [], 'entities': []}
-
    -
    -
    -
    -class PeopleEntity(body, dim=None, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, unit='', entity_type='people', *, range, type='value')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    A variant of numerical entity which addresses collections of people.

    -

    Example sentences that contain the entity are: -- “N people”, where N is a whole number. -- “A couple” ~ 2 people. -- “I have triplets” ~ 3 children.

    -
    -
    -entity_type: Optional[str]
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    PeopleEntity

    -
    -
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -345,14 +277,7 @@

    Workflow Integration diff --git a/docs/source/dialogy.types.entity.time.html b/docs/source/dialogy.types.entity.time.html index 5047dd32..fb269d75 100644 --- a/docs/source/dialogy.types.entity.time.html +++ b/docs/source/dialogy.types.entity.time.html @@ -6,7 +6,7 @@ - dialogy.types.entity.time package - dialogy 0.9.3 documentation + dialogy.types.entity.time package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,245 +220,8 @@

    dialogy.types.entity.time package

    -
    -

    Module contents

    -
    -

    Time Entity

    -

    Provides the entity class for representing time in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("tomorrow")
-Out[3]: [TimeEntity(body='tomorrow', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value='2022-02-16T00:00:00.000+05:30', entity_type='date', origin='value', dim='time', grain='day')]
-
-In [4]: duckling_plugin.parse("monday")
-Out[4]: [TimeEntity(body='monday', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value='2022-02-21T00:00:00.000+05:30', entity_type='date', origin='value', dim='time', grain='day')]
-
-In [5]: duckling_plugin.parse("27th jan")
-Out[5]: [TimeEntity(body='27th jan', type='value', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, value='2023-01-27T00:00:00.000+05:30', entity_type='date', origin='value', dim='time', grain='day')]
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: _, output = workflow.run(Input(utterances=["tomorrow", "monday", "27th jan"]))
-
-In [5]: output
-Out[5]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 8},
-   'body': 'tomorrow',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.3333333333333333,
-   'alternative_index': 0,
-   'value': '2022-02-16T00:00:00.000+05:30',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 0, 'end': 6},
-   'body': 'monday',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.3333333333333333,
-   'alternative_index': 1,
-   'value': '2022-02-21T00:00:00.000+05:30',
-   'entity_type': 'date',
-   'grain': 'day'},
-  {'range': {'start': 0, 'end': 8},
-   'body': '27th jan',
-   'type': 'value',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.3333333333333333,
-   'alternative_index': 2,
-   'value': '2023-01-27T00:00:00.000+05:30',
-   'entity_type': 'date',
-   'grain': 'day'}]}
-
    -
    -
    -
    -class TimeEntity(body, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, values=NOTHING, value=None, entity_type=None, origin='value', dim='time', grain=None, *, range, type='value')[source]
    -

    Bases: dialogy.types.entity.base_entity.BaseEntity

    -

    Entities that can be parsed to obtain date, time or datetime values.

    -

    Example sentences that contain the entity are: -- “I have a flight at 6 am.” -- “I have a flight at 6th December.” -- “I have a flight at 6 am today.”

    -

    Attributes: -- grain tells us the smallest unit of time in the utterance

    -
    -
    -collect_datetime_values()[source]
    -

    Collect all datetime values from the entity

    -
    -
    Returns
    -

    List of datetime values

    -
    -
    Return type
    -

    List[str]

    -
    -
    -
    -
    -
    -dim: str
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    TimeEntity

    -
    -
    -
    -
    -
    -get_value()[source]
    -

    Return the date string in ISO format from the dictionary passed

    -
    date = {
-    "value": "2021-04-17T16:00:00.000+05:30",
-    "grain": "hour",
-    "type": "value"
-}
-
    -
    -
    -
    Parameters
    -

    date (Dict[str, str]) – Dictionary which stores the datetime in ISO format, grain and type

    -
    -
    Returns
    -

    date["value"]

    -
    -
    Return type
    -

    Optional[datetime]

    -
    -
    -
    -
    -
    -grain: str
    -
    -
    -
    -is_uniq_date_from_values()[source]
    -

    Check a list has a unique day

    -

    Where day is the date number for a month.

    -
    -
    Returns
    -

    True if there is a unique day in all datetime values

    -
    -
    Return type
    -

    bool

    -
    -
    -
    -
    -
    -is_uniq_day_from_values()[source]
    -

    Check a list has a unique weekday

    -

    Where weekday ranges from 0 for Monday till 6 for Sunday -Ex: for “2021-04-17T16:00:00.000+05:30”, the weekday is 5 (Saturday)

    -

    Duckling may return 3 datetime values in chronological order. For cases where the utterance is “Monday”, -(or any weekday for that matter) we will get 3 values, each value a week apart. -We pair these dates and check the difference is 1 week long (7 days).

    -

    For example:

    -
    # Say, our 3 date values are:
-date values = [21-04-2021, 28-04-2021, 05-05-2021]
-
-# then we generate pairs as:
-[(21-04-2021, 28-04-2021), (28-04-2021, 05-05-2021)]
-
    -
    -

    If the difference between these dates is divisible by seven, then the input was a weekday.

    -
    -
    Returns
    -

    True if there is a unique weekday in all datetime values

    -
    -
    Return type
    -

    bool

    -
    -
    -
    -
    -
    -origin: str
    -
    -
    -
    -set_entity_type()[source]
    -

    Returns the type for this entity

    -

    To pinpoint a time for any action, we need both DATE (either explicit or inferred) and TIME. -We can get these in 3 possible manners:

    -
      -
    • -
      both DATE and TIME, the entity type here will be DATETIME
        -

      • Positive examples: tonight, tomorrow 4pm, 7th april 2019 6pm, last 3 hours (Because it tells us about the date as well, that is today)

        • -

      • Negative examples: Yesterday I went to park at 3pm (because the user did not say yesterday and 3pm together)

        • -
      -
      -
      -
      • -
    • -
      only DATE, the entity type here will be DATE
        -

      • Positive examples: last month, last year, last 3 months, last 6 months, tomorrow, 27th oct 1996, 1st October 2019 till yesterday, yesterday, today, tomorrow, October, 2019, 27th October 1996 to 29th October 1996 etc

        • -

      • Negative examples: month, year, 6 months, 3 months (Since these do not have a fixed start and end date, it could have been 6 months from today, 6 months from tomorrow, previous 6 months. Therefore it only tells us a duration.)

        • -
      -
      -
      -
      • -
    • -
      only TIME, the entity type here will be TIME
        -

      • Positive examples: 3pm, 5pm, night, morning, etc

        • -

      • Negative examples: tonight (because the date part today is in it), tomorrow night, 6th April 5pm, 7 hours (duration because they didn’t mention a start/end time. It could be 7 hours from any time) etc

        • -
      -
      -
      -
      • -
    -

    if we get the grain as one of (“day”, “week”, “month”, “quarter”, “year”), -then the entity type is DATE -if we get grain as one of (“hour”, “second”, “minute”) and we only have 1 value in self.values, -then the entity type is DATETIME. -if we get grain as one of (“hour”, “second”, “minute”) and we have either a unique date or a unique weekday, -then the entity type is DATETIME. -else the entity type is TIME.

    -
    -
    Returns
    -

    one of (date, time, datetime)

    -
    -
    Return type
    -

    str

    -
    -
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -514,14 +277,7 @@

    Workflow Integration diff --git a/docs/source/dialogy.types.entity.time_interval.html b/docs/source/dialogy.types.entity.time_interval.html index 6049b52b..8c8161e0 100644 --- a/docs/source/dialogy.types.entity.time_interval.html +++ b/docs/source/dialogy.types.entity.time_interval.html @@ -6,7 +6,7 @@ - dialogy.types.entity.time_interval package - dialogy 0.9.3 documentation + dialogy.types.entity.time_interval package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -220,173 +220,8 @@

    dialogy.types.entity.time_interval package

    -
    -

    Module contents

    -
    -

    Time Interval Entity

    -

    Provides the entity class for representing time intervals in natural language. This entity is obtained via DucklingPlugin.

    -
    -

    Plugin Walkthrough

    -
    In [1]: from dialogy.plugins import DucklingPlugin
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: duckling_plugin.parse("weekend")
-Out[3]: [TimeIntervalEntity(body='weekend', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, entity_type='datetime', origin='value', dim='time', grain='hour', type='interval', from_value='2022-02-18T18:00:00.000+05:30', to_value='2022-02-21T00:00:00.000+05:30', values=[{'to': {'value': '2022-02-21T00:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-02-18T18:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}, {'to': {'value': '2022-02-28T00:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-02-25T18:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}, {'to': {'value': '2022-03-07T00:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-03-04T18:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}], value={'from': '2022-02-18T18:00:00.000+05:30', 'to': '2022-02-21T00:00:00.000+05:30'})]
-
-In [4]: duckling_plugin.parse("between 4 to 5pm")
-Out[4]: [TimeIntervalEntity(body='between 4 to 5pm', parsers=['DucklingPlugin'], score=1.0, alternative_index=0, alternative_indices=[0], latent=False, entity_type='time', origin='value', dim='time', grain='hour', type='interval', from_value='2022-02-16T16:00:00.000+05:30', to_value='2022-02-16T18:00:00.000+05:30', values=[{'to': {'value': '2022-02-16T18:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-02-16T16:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}, {'to': {'value': '2022-02-17T18:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-02-17T16:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}, {'to': {'value': '2022-02-18T18:00:00.000+05:30', 'grain': 'hour'}, 'from': {'value': '2022-02-18T16:00:00.000+05:30', 'grain': 'hour'}, 'type': 'interval'}], value={'from': '2022-02-16T16:00:00.000+05:30', 'to': '2022-02-16T18:00:00.000+05:30'})]
-
    -
    -
    -
    -

    Workflow Integration

    -
    In [1]: from dialogy.base import Input
-   ...: from dialogy.plugins import DucklingPlugin
-   ...: from dialogy.workflow import Workflow
-   ...: 
-
-In [2]: duckling_plugin = DucklingPlugin(
-   ...:     dest="output.entities",
-   ...:     dimensions=["time"],
-   ...:     locale="en_IN",
-   ...:     timezone="Asia/Kolkata",
-   ...: )
-   ...: 
-
-In [3]: workflow = Workflow([duckling_plugin])
-
-In [4]: _, output = workflow.run(Input(utterances=["weekend", "between 4 to 5pm"]))
-
-In [5]: output
-Out[5]: 
-{'intents': [],
- 'entities': [{'range': {'start': 0, 'end': 7},
-   'body': 'weekend',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.5,
-   'alternative_index': 0,
-   'entity_type': 'datetime',
-   'grain': 'hour',
-   'type': 'interval',
-   'from_value': '2022-02-18T18:00:00.000+05:30',
-   'to_value': '2022-02-21T00:00:00.000+05:30',
-   'value': {'from': '2022-02-18T18:00:00.000+05:30',
-    'to': '2022-02-21T00:00:00.000+05:30'}},
-  {'range': {'start': 0, 'end': 16},
-   'body': 'between 4 to 5pm',
-   'parsers': ['DucklingPlugin'],
-   'score': 0.5,
-   'alternative_index': 1,
-   'entity_type': 'time',
-   'grain': 'hour',
-   'type': 'interval',
-   'from_value': '2022-02-16T16:00:00.000+05:30',
-   'to_value': '2022-02-16T18:00:00.000+05:30',
-   'value': {'from': '2022-02-16T16:00:00.000+05:30',
-    'to': '2022-02-16T18:00:00.000+05:30'}}]}
-
    -
    -
    -
    -class TimeIntervalEntity(body, parsers=NOTHING, score=None, alternative_index=None, alternative_indices=None, latent=False, entity_type=None, origin='value', dim='time', grain=None, type='value', from_value=None, to_value=None, *, range, values=None, value=None)[source]
    -

    Bases: dialogy.types.entity.time.TimeEntity

    -

    Entities that can be parsed to obtain date, time or datetime interval.

    -
      -

    • “I need a flight between 6 am to 10 am.”

      • -

    • “I have a flight at 6 am to 5 pm today.”

      • -
    -

    Attributes: -- origin -- value is a Dictionary which has either keys ‘from’ and ‘to’ or both

    -
    -
    -collect_datetime_values()[source]
    -

    Collect all datetime values from the entity

    -
    -
    Returns
    -

    List of datetime values

    -
    -
    Return type
    -

    List[str]

    -
    -
    -
    -
    -
    -dim: str = 'time'
    -
    -
    -
    -classmethod from_duckling(d, alternative_index)[source]
    -
    -
    Return type
    -

    TimeEntity

    -
    -
    -
    -
    -
    -from_value: Optional[datetime.datetime]
    -
    -
    -
    -get_value()[source]
    -

    Return the date string in ISO format from the dictionary passed

    -
    date = {
-    "from": {
-        "value": "2021-04-16T16:00:00.000+05:30",
-        "grain": "hour"
-    },
-    "type": "interval"
-}
-
    -
    -
    -
    Parameters
    -

    date (Dict[str, str]) – Dictionary which stores the datetime in ISO format, grain and type

    -
    -
    Returns
    -

    date["value"]

    -
    -
    Return type
    -

    Optional[datetime]

    -
    -
    -
    -
    -
    -origin: str = 'interval'
    -
    -
    -
    -to_value: Optional[datetime.datetime]
    -
    -
    -
    -type: str
    -
    -
    -
    -value: Dict[str, Any]
    -
    -
    -
    -value_keys = {'from', 'to', 'type'}
    -
    -
    -
    -values: List[Dict[str, Any]]
    -
    -
    -
    -
    +
    +

    Module contents

    @@ -442,14 +277,7 @@

    Workflow Integration diff --git a/docs/source/dialogy.types.html b/docs/source/dialogy.types.html index afb3115b..fb315309 100644 --- a/docs/source/dialogy.types.html +++ b/docs/source/dialogy.types.html @@ -6,7 +6,7 @@ - dialogy.types package - dialogy 0.9.3 documentation + dialogy.types package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -227,7 +227,7 @@

    Subpackagesdialogy.types.entity package

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,18 +222,16 @@

    dialogy.types.intent package

    Module contents

    -
    -

    Intent Classification

    -

    Automatic categorization of texts into known classes. When solving SLU problems -the objective is to answer What was meant? instead of “What was said?”.

    -

    For an example utterance “I lost my wallet” a bank’s support executives would -understand that they must follow the operating procedure for blocking all transaction -accounts. We want to train machines to have the ability to interpret these cues.

    -

    We call the classes of meaning as Intent.

    -

    We can obtain intents from XLMRMultiClass and MLPMultiClass plugin.

    +

    Type definition for Intent.

    +

    Import classes:

    +
    +
      +

    • Intent

      • +
    +
    -class Intent(*, name, score=0.0, alternative_index=None, parsers=NOTHING, slots=NOTHING)[source]
    +class Intent(name, score, parsers=None, alternative_index=0, slots=None)[source]

    Bases: object

    An instance of this class contains the name of the action associated with a body of text.

    @@ -254,54 +252,11 @@

    dialogy.types.intent package -
    -alternative_index: Optional[int]
    -

    Tells us the index of transcript that yeilds this intent. Since our featurizer -concatenates all transcripts, we currently don’t have a use for it.

    -

    If we were to predict the intent for each transcript separately and vote, in those -cases it would be helpful to log this property.

    -

    apply(rules)[source]
    -

    Create slots using rules.

    -

    An intent can hold different entities within associated slot-types. -We parse rules that describe this mapping and create placeholders within an intent for -holding entities found.

    -

    An example rule would look like:

    -
    rules = {
-    "wake_up_alarm": {          # name of the intent.
-        "datetime_slot": [      # "datetime_slot" is the only slot associated with this intent.
-            "date",             # Only "date", "time", "datetime" entities can fill the "datetime_slot"
-            "time",
-            "datetime"
-        ]
-    },
-    "ticket_booking": {         # name of the intent.
-        "origin": [             # A slot that will fill only location type entities.
-            "location"          # The entity types that can be filled in the slot.
-        ],
-        "destination": [        # Another slot that will fill only location type entities.
-            "location"
-        ],
-        "start_date": [
-            "date"
-        ],
-        "end_date": [
-            "date"
-        ],
-        "passengers": [
-            "people",
-            "number"
-        ],
-        "budget": [
-            "amount-of-money"
-        ]
-    }
-}
-
    -
    +

    Create slots using rules.

    +

    An intent can hold different entities within associated slot-types.

    Parameters

    rules (Rule) – A configuration for slot names and entity types associated with this instance of Intent.

    @@ -318,7 +273,7 @@

    dialogy.types.intent package cleanup()[source]

    Remove slots that were not filled.

    -
    +
    Return type

    None

    @@ -328,11 +283,11 @@

    dialogy.types.intent package fill_slot(entity, fill_multiple=False)[source]

    Update slots[slot_type].values with a single entity.

    -

    We will explore the possibility of sharing/understanding the meaning of multiple entities +

    We will explore the possibility of sharing/understanding the meaning of multiple entities in the same slot type.

    There maybe cases where we want to fill multiple entities of the same type within a slot. In these cases fill_multiple should be set to True.

    -

    The RuleBasedSlotFillerPlugin has a detailed demo for this.

    +

    The RuleBasedSlotFillerPlugin has a detailed demo for this.

    Parameters

    @@ -371,32 +326,8 @@

    dialogy.types.intent package -
    -name: str
    -

    The description of an intent separated by ‘_’

    -

    -
    -
    -parsers: List[str]
    -

    Contains a list of plugins that have created, updated this Intent.

    -
    -
    -
    -score: float
    -

    A positive real number that represents confidence of this intent -being the meaning of an utterance. Higher is better.

    -

    Should be a value between 0 and 1.

    -
    -
    -
    -slots: Dict[str, dialogy.types.slots.Slot]
    -

    Placeholders for entities that are relevant to this intent. -More details are available in the doc for slots.

    -
    - @@ -451,10 +382,7 @@

    dialogy.types.intent package diff --git a/docs/source/dialogy.types.plugin.html b/docs/source/dialogy.types.plugin.html index 0f523f17..943d0686 100644 --- a/docs/source/dialogy.types.plugin.html +++ b/docs/source/dialogy.types.plugin.html @@ -6,7 +6,7 @@ - dialogy.types.plugin package - dialogy 0.9.3 documentation + dialogy.types.plugin package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation diff --git a/docs/source/dialogy.types.signal.html b/docs/source/dialogy.types.signal.html index a9c55078..d5df4506 100644 --- a/docs/source/dialogy.types.signal.html +++ b/docs/source/dialogy.types.signal.html @@ -6,7 +6,7 @@ - dialogy.types.signal package - dialogy 0.9.3 documentation + dialogy.types.signal package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation diff --git a/docs/source/dialogy.types.slots.html b/docs/source/dialogy.types.slots.html index 695cbd4d..3be866f6 100644 --- a/docs/source/dialogy.types.slots.html +++ b/docs/source/dialogy.types.slots.html @@ -6,7 +6,7 @@ - dialogy.types.slots package - dialogy 0.9.3 documentation + dialogy.types.slots package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -222,13 +222,19 @@

    dialogy.types.slots package

    Module contents

    -

    Type definition for Slots. Read Intent and Slot Filling for details.

    +

    Type definition for Slots.

    +

    Import classes:

    +
    +
      +

    • Slot

      • +
    +
    -class Slot(*, name, types=NOTHING, values=NOTHING)[source]
    +class Slot(name, types, values)[source]

    Bases: object

    Slot Type

    -

    Keys are: +

    Keys are: - name of the slot - type is a list of types that can fill this slot - values list of entities extracted

    @@ -266,18 +272,6 @@

    dialogy.types.slots package -
    -name: str
    -

    -
    -
    -types: List[str]
    -
    -
    -
    -values: List[dialogy.types.entity.base_entity.BaseEntity]
    -
    diff --git a/docs/source/dialogy.types.utterances.html b/docs/source/dialogy.types.utterances.html index 2ef4e127..5615658f 100644 --- a/docs/source/dialogy.types.utterances.html +++ b/docs/source/dialogy.types.utterances.html @@ -6,7 +6,7 @@ - dialogy.types.utterances package - dialogy 0.9.3 documentation + dialogy.types.utterances package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation diff --git a/docs/source/dialogy.utils.html b/docs/source/dialogy.utils.html index 4d1f57c6..e4f9899f 100644 --- a/docs/source/dialogy.utils.html +++ b/docs/source/dialogy.utils.html @@ -6,7 +6,7 @@ - dialogy.utils package - dialogy 0.9.3 documentation + dialogy.utils package - dialogy 0.9.4 documentation @@ -125,7 +125,7 @@
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -223,73 +223,8 @@

    dialogy.utils package

    Submodules

    -
    -

    dialogy.utils.datetime module

    -
    -
    -dt2timestamp(date_time)[source]
    -

    Converts a python datetime object to unix-timestamp.

    -
    -
    Parameters
    -

    date_time (datetime) – An instance of datetime.

    -
    -
    Returns
    -

    Unix timestamp integer.

    -
    -
    Return type
    -

    int

    -
    -
    -
    -
    -
    -is_unix_ts(ts)[source]
    -

    Check if the input is a unix timestamp.

    -
    -
    Parameters
    -

    ts (int) – A unix timestamp (13-digit).

    -
    -
    Returns
    -

    True if ts is a unix timestamp, else False.

    -
    -
    Return type
    -

    bool

    -
    -
    -
    -
    -
    -make_unix_ts(tz='UTC')[source]
    -

    Convert date in ISO 8601 format to unix ms timestamp.

    -
    In [1]: from dialogy.utils.datetime import make_unix_ts
-
-In [2]: ts = make_unix_ts("Asia/Kolkata")("2022-02-07T19:39:39.537827")
-
-In [3]: ts == 1644241599537
-Out[3]: True
-
    -
    -
    -
    Parameters
    -

    tz (Optional[str], optional) – A timezone string, defaults to “UTC”

    -
    -
    Returns
    -

    A callable that converts a date in ISO 8601 format to unix ms timestamp.

    -
    -
    Return type
    -

    Callable[[str], int]

    -
    -
    -
    -
    -
    -unix_ts_to_datetime(reference_time, timezone='UTC')[source]
    -
    -
    Return type
    -

    datetime

    -
    -
    -
    +
    +

    dialogy.utils.datetime module

    dialogy.utils.file_handler module

    @@ -375,6 +310,12 @@

    Submodules

    dialogy.utils.logger module

    Module provides access to logger.

    This needs to be used sparingly, prefer to raise specific exceptions instead.

    +
    +
    Import functions:
      +

    • change_log_level

      • +
    +
    +

    dialogy.utils.misc module

    @@ -387,6 +328,22 @@

    Submodules +
    +dt2timestamp(date_time)[source]
    +

    Converts a python datetime object to unix-timestamp.

    +
    +
    Parameters
    +

    date_time (datetime) – An instance of datetime.

    +
    +
    Returns
    +

    Unix timestamp integer.

    +
    +
    Return type
    +

    int

    +
    +
    +
    +
    traverse_dict(obj, properties)[source]

    Traverse a dictionary for a given list of properties.

    @@ -416,36 +373,22 @@

    SubmodulesIn [4]: traverse_dict(input_, ["planets", "mars", 0 , "languages", 3, "bop"]) --------------------------------------------------------------------------- IndexError Traceback (most recent call last) -Input In [4], in <module> +<ipython-input-4-556a11057179> in <module> ----> 1 traverse_dict(input_, ["planets", "mars", 0 , "languages", 3, "bop"]) -File ~/Programs/pythoncode/dialogy/dialogy/utils/misc.py:52, in traverse_dict(obj, properties) - 13 """ - 14 Traverse a dictionary for a given list of properties. - 15 - (...) - 49 :raises TypeError: Properties don't describe a path due to possible type error. - 50 """ - 51 try: ----> 52 return reduce(lambda o, k: o[k], properties, obj) - 53 except KeyError as key_error: - 54 raise KeyError( - 55 f"Missing property {key_error} in {obj}. Check the types. Failed for path {properties}" - 56 ) from key_error +~/.pyenv/versions/3.8.7/envs/venv_dialogy/lib/python3.8/site-packages/dialogy/utils/misc.py in traverse_dict(obj, properties) + 51 """ + 52 try: +---> 53 return reduce(lambda o, k: o[k], properties, obj) + 54 except KeyError as key_error: + 55 raise KeyError( -File ~/Programs/pythoncode/dialogy/dialogy/utils/misc.py:52, in traverse_dict.<locals>.<lambda>(o, k) - 13 """ - 14 Traverse a dictionary for a given list of properties. - 15 - (...) - 49 :raises TypeError: Properties don't describe a path due to possible type error. - 50 """ - 51 try: ----> 52 return reduce(lambda o, k: o[k], properties, obj) - 53 except KeyError as key_error: - 54 raise KeyError( - 55 f"Missing property {key_error} in {obj}. Check the types. Failed for path {properties}" - 56 ) from key_error +~/.pyenv/versions/3.8.7/envs/venv_dialogy/lib/python3.8/site-packages/dialogy/utils/misc.py in <lambda>(o, k) + 51 """ + 52 try: +---> 53 return reduce(lambda o, k: o[k], properties, obj) + 54 except KeyError as key_error: + 55 raise KeyError( IndexError: list index out of range

    @@ -692,33 +635,22 @@

    Submodulesnormalize(maybe_utterance, key='transcript')[source]

    Adapt various non-standard ASR alternative forms.

    The output will be a list of strings since models will expect that.

    -
    In [1]: In [1]: from dialogy.utils.normalize_utterance import normalize
-   ...: In [2]: # A popular case
-   ...: In [3]: normalize([[{"transcript": "this"}, {"transcript": "works"}]])
-   ...: 
-Out[1]: ['this', 'works']
-
    -
    -
    In [2]: In [3]: # A case with multiple utterances
-   ...: In [4]: normalize([
-   ...:             [{"transcript": "hello hello?", "transcript": "yellow yellow?"}],
-   ...:             [{"transcript": "I wanted to check"}],
-   ...:             [{"transcript": "if you have space for us?"}]
-   ...:         ])
-   ...: 
-Out[2]: ['yellow yellow? I wanted to check if you have space for us?']
-
    -
    -
    In [3]: In [5]: normalize([{"transcript": "I wanted to know umm hello?"}])
-Out[3]: ['I wanted to know umm hello?']
-
    -
    -
    In [4]: In [6]: normalize(["I wanted to know umm hello?"])
-Out[4]: ['I wanted to know umm hello?']
-
    -
    -
    In [5]: In [7]: normalize("I wanted to know umm hello?")
-Out[5]: ['I wanted to know umm hello?']
+
    In [1]: from dialogy.utils.normalize_utterance import normalize
+
+In [2]: normalize([[{"transcript": "hi"}]])
+Out[2]: ['hi']
+
+In [3]: normalize([[{"transcript": "hello"}], [{"transcript": "world"}]])
+Out[3]: ['hello world']
+
+In [4]: normalize([{"transcript": "I wanted to know umm hello?"}])
+Out[4]: ['i wanted to know umm hello?']
+
+In [5]: normalize(["I wanted to know umm hello?"])
+Out[5]: ['i wanted to know umm hello?']
+
+In [6]: normalize("I wanted to know umm hello?")
+Out[6]: ['i wanted to know umm hello?']
 
    
 
    
 
    
@@ -800,7 +732,7 @@ 
    Submodulesdialogy.utils package
    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -225,121 +225,70 @@

    Submodules

    dialogy.workflow.workflow module

    -

    The Workflow represents a black box for an SLU microservice, and its terminals, input and output -represent request response structure of the SLU API.

    -
    -

    Note

    -

    The Input has some extra properties that aren’t part of the SLU API. -We have kept reserves for intermediates, and flattened as most other nested inputs -aren’t required and add to bloat.

    -
    -
    -

    Attention

    -

    Input and Output instances are immutable.

    -
    -
    -

    What does it do?

    -

    Produce the response for a single turn of a conversation. It recieves (not an exhaustive list):

    +

    A workflow is supposed to run tasks that can be anticipated well in advance.

    +

    Here are few demo’s where you can see mock workflow in action.

    -

    as inputs, parses these, produces a few intermediates and finally returns the intent -and optionally entities. Details will be documented within the specific plugins.

    -
    - classDiagram - direction TB - Workflow --> "1" Input: has - Workflow --> "1" Output: has - Workflow --> "many" Plugin: has - - class Workflow { - Workflow: +Input input - Workflow: +Output output - Workflow: +set(path: str, value: Any) - Workflow: +run(input: Input) - } - - class Input { - +List~Utterance~ utterance - +int reference_time - +str current_state - +str lang - +str locale - +dict slot_tracker - +str timezone - +dict json() - } - - class Output { - +List~Intent~ intents - +List~BaseEntity~ entities - +dict json() - } - - <<abstract>> Plugin - class Plugin { - +str dest - *any utility(input: Input, output: Output) - } -
    -
    -

    How does it do it?

    +

    A workflow is a hollow conduit, think of a vertically hanging pipe without any medium. If you were to drop a block of … +anything? through it, it would pass through with a thud on the ground (yes we assumed gravity).

    +
    +

    Structure

    +

    A workflow allows flexibility, that’s why. There is very little structure to it. We have:

      -

    • A workflow contains a sequence of plugins. The sequence is important.

      • -

    • The sequence describes the order in which plugins are run.

      • -

    • The plugins can save their output within the workflow’s input or output.

      • -

    • After execution of all plugins, the workflow returns a pair of serialized input and output.

      • +

    • input

      • +

    • output

      • +

    • plugins

    -
    - flowchart TB - - subgraph Workflow - - subgraph input - utterance - reference_time - end - - subgraph output - intents - entities - end - - input --> plugin - output --> plugin - plugin -->|update| input - plugin -->|update| output - plugin -->|foreach| plugin - - end - output --> output_json - input --> input_json - - output_json --> workflow_output - input_json --> workflow_output - - subgraph Response - output_json - end -
    +

    Apart from these, we expect at the core, an inference function with machine learning models. Which ones? A N Y ones. +As long as you have the compute, there is no restriction. Use statistical models or DL or a bunch of conditions, +your Workflow won’t judge you.

    +
    +
    +

    Advantages

    +

    The Plugin concept takes care of the sauciness 🍅 of this project. Any functionality can be bundled into +a Plugin and they are portable over to foreign workflows. A Plugin proxies inputs through +an access function (an argument to every Plugin) and relays output through a mutate +function (another argument for every Plugin). These two functions define interactions between many many +(sic) Plugins without knowing the inner workings of a Workflow.

    +

    Take a look at DucklingPlugin, this plugin handles inputs, manages the default json output +into neatly bundled BaseEntity and other similar classes. Another plugin +RuleBasedSlotFillerPlugin takes care of slot names and the entity types that should be filled +within.

    +

    If your classifier predicts an Intent with Slots supporting any of those entities, then +slot-filling is not a worry.

    +

    The aim of this project is to be largest supplier of plugins for SLU applications.

    +
    +

    Warning

    +

    The Workflow class is not supposed to be used as it is. Ideally it should have been an +abstract class. There are some design considerations which make that a bad choice. We want methods to be overridden +to offer flexibility of use.

    +
    +
    -class Workflow(plugins=NOTHING, debug=False, *, input=None, output=None)[source]
    +class Workflow(plugins=NOTHING, debug=False, *, input=NOTHING, output=NOTHING)[source]

    Bases: object

    -

    SLU API blackbox.

    -
    +

    This is a light but fairly flexible workflow for building a machine learning pipeline.

    +

    Requirements +- A list of pre-processing functions. +- A list of post-processing functions.

    +

    Abstract classes put constraints on method signatures which isn’t desired because a couple of methods +here could use more arguments, say, load_model() requires path and version and in some other cases +path, version and language.

    +
    +
    Parameters
    +
      +

    • plugins (Optional[List[PluginFn]]) – A list of functions to execute before inference.

      • +

    • debug (bool) – log level shifts to debug if True.

      • +
    +
    +
    +
    NON_SERIALIZABLE_FIELDS = ['plugins', 'debug', 'lock']
    -
    -
    -debug
    -

    Switch on/off the debug logs for the workflow.

    -
    execute()[source]
    @@ -347,9 +296,15 @@

    How does it do it? +
    Args:

    processors (List): The list of preprocess or postprocess functions.

    +
    +
    Raises:

    TypeError: If any element in processors list is not a Callable.

    +
    +

    Return type
    -

    Workflow

    +

    None

    @@ -359,16 +314,14 @@

    How does it do it?workflow.input and workflow.output.

    Return type
    -

    Tuple[Dict[str, Any], Dict[str, Any]]

    +

    None

    -input: Optional[dialogy.base.input.Input]
    -

    Represents the SLU API request object. A few nested properties and intermediates are flattened out -for readability.

    -
    +input: Dict[str, Any] +
    json()[source]
    @@ -382,49 +335,37 @@

    How does it do it?
    lock: _thread.allocate_lock
    -

    A workflow isn’t thread-safe by itself. We need locks -to prevent race conditions.

    -

    +
    -output: Optional[dialogy.base.output.Output]
    -

    Represents the SLU API response object.

    -
    -
    -
    -plugins
    -

    List of plugins.

    -
    +output: Dict[str, Any] +
    run(input_)[source]

    Get final results from the workflow.

    The current workflow exhibits the following simple procedure: pre-processing -> inference -> post-processing.

    +
    +
    Args:

    input_ (Any): This function receives any arbitrary input. Subclasses may enforce +a stronger check.

    +
    +
    Returns:

    (Any): This function can return any arbitrary value. Subclasses may enforce a stronger check.

    +
    +
    Return type
    -

    Tuple[Dict[str, Any], Dict[str, Any]]

    +

    Any

    -
    -set(path, value, replace=False)[source]
    -

    Set attribute path with value.

    -

    This method (re)-sets the input or output object without losing information -from previous instances.

    +
    +set_io()[source]
    +

    Use this method to keep workflow-io in the same format as expected.

    -
    Parameters
    -
      -

    • path (str) – A ‘.’ separated attribute path.

      • -

    • value (Any) – A value to set.

      • -
    -
    -
    Returns
    -

    This instance

    -
    Return type
    -

    Workflow

    +

    None

    @@ -440,7 +381,7 @@

    How does it do it?

    training_data (pd.DataFrame) – [description]

    Return type
    -

    Workflow

    +

    None

    @@ -506,8 +447,8 @@

    How does it do it?dialogy.workflow package

    -
    dialogy 0.9.3 documentation
     +
    dialogy 0.9.4 documentation
    @@ -148,7 +148,7 @@
    - dialogy 0.9.3 documentation + dialogy 0.9.4 documentation @@ -252,7 +252,7 @@

    dialogydialogy.utils package
    • Submodules
      • -
    • dialogy.utils.datetime module
      • +
    • dialogy.utils.datetime module
    • dialogy.utils.file_handler module
    • dialogy.utils.logger module
    • dialogy.utils.misc module
      • diff --git a/tests/plugin/text/slot_filler/test_rule_slot_filler.py b/tests/plugin/text/slot_filler/test_rule_slot_filler.py index 5b7890ae..c4003cfc 100644 --- a/tests/plugin/text/slot_filler/test_rule_slot_filler.py +++ b/tests/plugin/text/slot_filler/test_rule_slot_filler.py @@ -180,12 +180,8 @@ def access(workflow: Workflow) -> Any: # `workflow.output[0]` is the `Intent` we created. # The `entity_1_slot` and `entity_2_slot` are filled. - assert output[const.INTENTS][0]["slots"][0]["values"] == [ - entity_1.json() - ] - assert output[const.INTENTS][0]["slots"][1]["values"] == [ - entity_2.json() - ] + assert output[const.INTENTS][0]["slots"][0]["values"] == [entity_1.json()] + assert output[const.INTENTS][0]["slots"][1]["values"] == [entity_2.json()] def test_slot_filling_multiple() -> None: @@ -232,12 +228,8 @@ def test_slot_filling_multiple() -> None: # `workflow.output[0]` is the `Intent` we created. # The `entity_1_slot` and `entity_2_slot` are filled. - assert output[const.INTENTS][0]["slots"][0]["values"] == [ - entity_1.json() - ] - assert output[const.INTENTS][0]["slots"][1]["values"] == [ - entity_2.json() - ] + assert output[const.INTENTS][0]["slots"][0]["values"] == [entity_1.json()] + assert output[const.INTENTS][0]["slots"][1]["values"] == [entity_2.json()] def test_slot_competition_fill_multiple() -> None: