added _Solution versions of SLUHandsOn up to Solution4

CNTK.sln

@@ -1160,14 +1160,19 @@ Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "ImageHandsOn", "ImageHandsO
 		Tutorials\ImageHandsOn\cifar10.pretrained.cmf = Tutorials\ImageHandsOn\cifar10.pretrained.cmf
 		Tutorials\ImageHandsOn\CifarConverter.py = Tutorials\ImageHandsOn\CifarConverter.py
 		Tutorials\ImageHandsOn\ImageHandsOn.cntk = Tutorials\ImageHandsOn\ImageHandsOn.cntk
+		Tutorials\SLUHandsOn\slu.forward.backward.cmf = Tutorials\SLUHandsOn\slu.forward.backward.cmf
 	EndProjectSection
 EndProject
 Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "SLUHandsOn", "SLUHandsOn", "{CC143D08-567D-4DAC-9E14-264749C19039}"
 	ProjectSection(SolutionItems) = preProject
 		Tutorials\SLUHandsOn\atis.test.ctf = Tutorials\SLUHandsOn\atis.test.ctf
 		Tutorials\SLUHandsOn\atis.train.ctf = Tutorials\SLUHandsOn\atis.train.ctf
+		Tutorials\SLUHandsOn\slu.forward.cmf = Tutorials\SLUHandsOn\slu.forward.cmf
+		Tutorials\SLUHandsOn\slu.forward.lookahead.cmf = Tutorials\SLUHandsOn\slu.forward.lookahead.cmf
 		Tutorials\SLUHandsOn\slu.forward.nobn.cmf = Tutorials\SLUHandsOn\slu.forward.nobn.cmf
 		Tutorials\SLUHandsOn\SLUHandsOn.cntk = Tutorials\SLUHandsOn\SLUHandsOn.cntk
+		Tutorials\SLUHandsOn\SLUHandsOn_Solution1.cntk = Tutorials\SLUHandsOn\SLUHandsOn_Solution1.cntk
+		Tutorials\SLUHandsOn\SLUHandsOn_Solution2.cntk = Tutorials\SLUHandsOn\SLUHandsOn_Solution2.cntk
 	EndProjectSection
 EndProject
 Global

Tutorials/SLUHandsOn/SLUHandsOn_Solution1.cntk

@@ -0,0 +1,86 @@
+# CNTK Configuration File for creating a slot tagger and an intent tagger.
+
+command = TrainTagger:TestTagger
+
+makeMode = false ; traceLevel = 0 ; deviceId = "auto"
+
+rootDir = "." ; dataDir  = "$rootDir$" ; modelDir = "$rootDir$/Models"
+
+modelPath = "$modelDir$/slu.cmf"
+
+vocabSize = 943 ; numLabels = 129 ; numIntents = 26    # number of words in vocab, slot labels, and intent labels
+
+# The command to train the LSTM model
+TrainTagger = {
+    action = "train"
+    BrainScriptNetworkBuilder = {
+        inputDim = $vocabSize$
+        labelDim = $numLabels$
+        embDim = 150
+        hiddenDim = 300
+
+        model = Sequential (
+            EmbeddingLayer {embDim} :                            # embedding
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :  ##### added
+            RecurrentLSTMLayer {hiddenDim, goBackwards=false} :  # LSTM
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :  ##### added
+            DenseLayer {labelDim, initValueScale=7}              # output layer
+        )
+
+        # features
+        query      = Input {inputDim}
+        slotLabels = Input {labelDim}
+
+        # model application
+        z = model (query)
+
+        # loss and metric
+        ce   = CrossEntropyWithSoftmax (slotLabels, z)
+        errs = ErrorPrediction         (slotLabels, z)
+
+        featureNodes    = (query)
+        labelNodes      = (slotLabels)
+        criterionNodes  = (ce)
+        evaluationNodes = (errs)
+        outputNodes     = (z)
+    }
+
+    SGD = {
+        maxEpochs = 8 ; epochSize = 36000
+
+        minibatchSize = 70
+
+        learningRatesPerSample = 0.01*2:0.005*12:0.001
+        gradUpdateType = "FSAdaGrad"
+        gradientClippingWithTruncation = true ; clippingThresholdPerSample = 15.0
+
+        firstMBsToShowResult = 10 ; numMBsToShowResult = 100
+    }
+
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.train.ctf"
+        randomize = true
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}
+
+# Test the model's accuracy (as an error count)
+TestTagger = {
+    action = "eval"
+    modelPath = $modelPath$
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.test.ctf"
+        randomize = false
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}

Tutorials/SLUHandsOn/SLUHandsOn_Solution2.cntk

@@ -0,0 +1,89 @@
+# CNTK Configuration File for creating a slot tagger and an intent tagger.
+
+command = TrainTagger:TestTagger
+
+makeMode = false ; traceLevel = 0 ; deviceId = "auto"
+
+rootDir = "." ; dataDir  = "$rootDir$" ; modelDir = "$rootDir$/Models"
+
+modelPath = "$modelDir$/slu.cmf"
+
+vocabSize = 943 ; numLabels = 129 ; numIntents = 26    # number of words in vocab, slot labels, and intent labels
+
+# The command to train the LSTM model
+TrainTagger = {
+    action = "train"
+    BrainScriptNetworkBuilder = {
+        inputDim = $vocabSize$
+        labelDim = $numLabels$
+        embDim = 150
+        hiddenDim = 300
+
+        LookaheadLayer (x) = Splice (x : FutureValue (0, x, defaultHiddenActivation=0))
+
+        model = Sequential (
+            EmbeddingLayer {embDim} :                            # embedding
+            LookaheadLayer :                   ##### added
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :
+            RecurrentLSTMLayer {hiddenDim, goBackwards=false} :  # LSTM
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :
+            DenseLayer {labelDim, initValueScale=7}              # output layer
+        )
+
+        # features
+        query      = Input {inputDim}
+        slotLabels = Input {labelDim}
+
+        # model application
+        z = model (query)
+
+        # loss and metric
+        ce   = CrossEntropyWithSoftmax (slotLabels, z)
+        errs = ErrorPrediction         (slotLabels, z)
+
+        featureNodes    = (query)
+        labelNodes      = (slotLabels)
+        criterionNodes  = (ce)
+        evaluationNodes = (errs)
+        outputNodes     = (z)
+    }
+
+    SGD = {
+        maxEpochs = 8 ; epochSize = 36000
+
+        minibatchSize = 70
+
+        learningRatesPerSample = 0.01*2:0.005*12:0.001
+        gradUpdateType = "FSAdaGrad"
+        gradientClippingWithTruncation = true ; clippingThresholdPerSample = 15.0
+
+        firstMBsToShowResult = 10 ; numMBsToShowResult = 100
+    }
+
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.train.ctf"
+        randomize = true
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}
+
+# Test the model's accuracy (as an error count)
+TestTagger = {
+    action = "eval"
+    modelPath = $modelPath$
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.test.ctf"
+        randomize = false
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}

Tutorials/SLUHandsOn/SLUHandsOn_Solution3.cntk

@@ -0,0 +1,93 @@
+# CNTK Configuration File for creating a slot tagger and an intent tagger.
+
+command = TrainTagger:TestTagger
+
+makeMode = false ; traceLevel = 0 ; deviceId = "auto"
+
+rootDir = "." ; dataDir  = "$rootDir$" ; modelDir = "$rootDir$/Models"
+
+modelPath = "$modelDir$/slu.cmf"
+
+vocabSize = 943 ; numLabels = 129 ; numIntents = 26    # number of words in vocab, slot labels, and intent labels
+
+# The command to train the LSTM model
+TrainTagger = {
+    action = "train"
+    BrainScriptNetworkBuilder = {
+        inputDim = $vocabSize$
+        labelDim = $numLabels$
+        embDim = 150
+        #hiddenDim = 300
+        hiddenDim = 150
+
+        BiRecurrentLSTMLayer {outDim} = {
+            F = RecurrentLSTMLayer {outDim, goBackwards=false}
+            G = RecurrentLSTMLayer {outDim, goBackwards=true}
+            apply (x) = Splice (F(x):G(x))
+        }.apply
+
+        model = Sequential (
+            EmbeddingLayer {embDim} :
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :
+            BiRecurrentLSTMLayer {hiddenDim} :
+            BatchNormalizationLayer {normalizationTimeConstant=2048} :
+            DenseLayer {labelDim, initValueScale=7}
+        )
+
+        # features
+        query      = Input {inputDim}
+        slotLabels = Input {labelDim}
+
+        # model application
+        z = model (query)
+
+        # loss and metric
+        ce   = CrossEntropyWithSoftmax (slotLabels, z)
+        errs = ErrorPrediction         (slotLabels, z)
+
+        featureNodes    = (query)
+        labelNodes      = (slotLabels)
+        criterionNodes  = (ce)
+        evaluationNodes = (errs)
+        outputNodes     = (z)
+    }
+
+    SGD = {
+        maxEpochs = 8 ; epochSize = 36000
+
+        minibatchSize = 70
+
+        learningRatesPerSample = 0.01*2:0.005*12:0.001
+        gradUpdateType = "FSAdaGrad"
+        gradientClippingWithTruncation = true ; clippingThresholdPerSample = 15.0
+
+        firstMBsToShowResult = 10 ; numMBsToShowResult = 100
+    }
+
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.train.ctf"
+        randomize = true
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}
+
+# Test the model's accuracy (as an error count)
+TestTagger = {
+    action = "eval"
+    modelPath = $modelPath$
+    reader = {
+        readerType = "CNTKTextFormatReader"
+        file = "$DataDir$/atis.test.ctf"
+        randomize = false
+        input = {
+            query        = { alias = "S0" ; dim = $vocabSize$ ;  format = "sparse" }
+            intentLabels = { alias = "S1" ; dim = $numIntents$ ; format = "sparse" }
+            slotLabels   = { alias = "S2" ; dim = $numLabels$ ;  format = "sparse" }
+        }
+    }
+}