Component calculating BLEU

configs/default/components/publishers/bleu_statistics.yml

@@ -0,0 +1,53 @@
+# This file defines the default values for the Accuracy statistics.
+
+####################################################################
+# 1. CONFIGURATION PARAMETERS that will be LOADED by the component.
+####################################################################
+
+# Flag indicating whether prediction are represented as distributions or indices (LOADED)
+# Options: True (expects distribution for each preditions)
+#          False (expects indices (max args))
+use_prediction_distributions: True
+
+# When set to True, performs masking of selected samples from batch (LOADED)
+# TODO!
+#use_masking: False
+
+# Weights of n-grams used when calculating the score.
+weights: [0.25, 0.25, 0.25, 0.25]
+
+streams: 
+  ####################################################################
+  # 2. Keymappings associated with INPUT and OUTPUT streams.
+  ####################################################################
+
+  # Stream containing targets (label ids) (INPUT)
+  targets: targets
+
+  # Stream containing batch of predictions (INPUT)
+  predictions: predictions
+
+  # Stream containing masks used for masking of selected samples from batch (INPUT)
+  #masks: masks
+
+globals:
+  ####################################################################
+  # 3. Keymappings of variables that will be RETRIEVED from GLOBALS.
+  ####################################################################
+
+  # Word mappings used for mappings of predictions/targets into list of words (RERIEVED)
+  word_mappings: word_mappings
+
+  ####################################################################
+  # 4. Keymappings associated with GLOBAL variables that will be SET.
+  ####################################################################
+
+statistics:
+  ####################################################################
+  # 5. Keymappings associated with statistics that will be ADDED.
+  ####################################################################
+
+  # Name used for collected statistics (ADDED).
+  bleu: bleu
+
+

configs/default/components/text/sentence_indexer.yml

@@ -25,6 +25,14 @@ import_word_mappings_from_globals: False
 # Flag informing whether word mappings will be exported to globals (LOADED)
 export_word_mappings_to_globals: False
 
+# Operation mode. If 'reverse' is True, then it will change indices into words (LOADED)
+reverse: False
+
+# Flag indicating whether inputs are represented as distributions or indices (LOADED)
+# Options: True (expects distribution for each input item in sequence)
+#          False (expects indices (max args))
+use_input_distributions: False
+
 streams: 
   ####################################################################
   # 2. Keymappings associated with INPUT and OUTPUT streams.

configs/wikitext/wikitext_language_modeling_rnn.yml

@@ -5,7 +5,7 @@ training:
     data_folder: &data_folder ~/data/language_modeling/wikitext-2
     dataset: &dataset wikitext-2
     subset: train
-    sentence_length: 50
+    sentence_length: 10
     batch_size:  64
 
   # optimizer parameters:
@@ -27,7 +27,7 @@ validation:
     data_folder: *data_folder
     dataset: *dataset
     subset: valid
-    sentence_length: 50
+    sentence_length: 20
     batch_size:  64
 
 # Testing parameters:
@@ -46,7 +46,7 @@ pipeline:
   # Source encoding - model 1.
   source_sentence_embedding:
     type: SentenceEmbeddings
-    priority: 1.1
+    priority: 1
     embeddings_size: 50
     pretrained_embeddings: glove.6B.50d.txt
     data_folder: *data_folder
@@ -61,7 +61,7 @@ pipeline:
   # Target encoding.
   target_indexer:
     type: SentenceIndexer
-    priority: 2.1
+    priority: 2
     data_folder: *data_folder
     import_word_mappings_from_globals: True
     streams:
@@ -86,12 +86,44 @@ pipeline:
     num_targets_dims: 2
     streams:
       targets: indexed_targets
-      loss: loss
 
+  # Prediction decoding.
+  prediction_decoder:
+    type: SentenceIndexer
+    priority: 10
+    # Reverse mode.
+    reverse: True
+    # Use distributions as inputs.
+    use_input_distributions: True
+    data_folder: *data_folder
+    import_word_mappings_from_globals: True
+    streams:
+      inputs: predictions
+      outputs: prediction_sentences
+
+
+  # Statistics.
+  batch_size:
+    type: BatchSizeStatistics
+    priority: 100.0
+
+  #accuracy:
+  #  type: AccuracyStatistics
+  #  priority: 100.1
+  #  streams:
+  #    targets: indexed_targets
+
+  bleu:
+    type: BLEUStatistics
+    priority: 100.2
+    streams:
+      targets: indexed_targets
+
+
   # Viewers.
   viewer:
     type: StreamViewer
-    priority: 100.1
-    input_streams: sources,indexed_targets,targets,predictions
+    priority: 100.3
+    input_streams: sources,targets,indexed_targets,prediction_sentences
 
 #: pipeline

ptp/components/mixins/word_mappings.py

@@ -17,31 +17,26 @@
 import os
 
 import ptp.components.utils.word_mappings as wm
-from ptp.components.component import Component
 
 
-class WordMappings(Component):
+class WordMappings(object):
     """
     Mixin class that handles the initialization of (word:index) mappings.
+    Assumes that it is mixed-in into class that is derived from the component.
+    .. warning::
+        Constructor (__init__) of the Component class has to be called before component of the mixin WordMapping class.
+
     """
-    def __init__(self, name, class_type, config):
+    def __init__(self): #, name, class_type, config):
         """
         Initializes the (word:index) mappings.
 
-        Loads parameters from configuration, 
-
-        :param name: Component name (read from configuration file).
-        :type name: str
-
-        :param class_type: Class type of the component (derrived from this class).
-
-        :param config: Dictionary of parameters (read from the configuration ``.yaml`` file).
-        :type config: :py:class:`ptp.configuration.ConfigInterface`
+        Assumes that Component was initialized in advance, which means that the self object possesses the following objects:
+            - self.config
+            - self.globals
+            - self.logger
 
         """
-        # Call constructors of parent classes.
-        Component.__init__(self, name, class_type, config)
-
         # Read the actual configuration.
         self.data_folder = os.path.expanduser(self.config['data_folder'])
 

ptp/components/models/sentence_embeddings.py

@@ -50,7 +50,7 @@ def __init__(self, name, config):
         """
         # Call base class constructors.
         Model.__init__(self, name, SentenceEmbeddings, config)
-        WordMappings.__init__(self, name, SentenceEmbeddings, config)
+        WordMappings.__init__(self)
 
         # Set key mappings.
         self.key_inputs = self.stream_keys["inputs"]

ptp/components/publishers/__init__.py

@@ -1,11 +1,13 @@
 from .accuracy_statistics import AccuracyStatistics
 from .batch_size_statistics import BatchSizeStatistics
+from .bleu_statistics import BLEUStatistics
 from .global_variable_publisher import GlobalVariablePublisher
 from .precision_recall_statistics import PrecisionRecallStatistics
 
 __all__ = [
     'AccuracyStatistics',
     'BatchSizeStatistics',
+    'BLEUStatistics',
     'GlobalVariablePublisher',
     'PrecisionRecallStatistics',
     ]