# Parameters can be overwritten on the command line
# for example: cntk configFile=myConfigFile RootDir=../.. 
# For running from Visual Studio add
# currentDirectory=$(SolutionDir)/<path to corresponding data folder> 

#command = trainNetwork:testNetwork:writeResults
command = trainNetwork:testNetwork

precision = "float"; traceLevel = 1 ; deviceId = "auto"

rootDir = ".." ; dataDir = "$rootDir$/DataSets/MNIST" ;
outputDir = "./Output" ;

modelPath = "$outputDir$/Models/07_Deconvolution_BS.model"
#stderr = "$outputDir$/07_Deconvolution_BS_out.txt"
#makemode=false

# TRAINING CONFIG
trainNetwork = {
    action = "train"
    
    BrainScriptNetworkBuilder = {
        cMap = 1
        model = inputFeatures => {
            conv1   = ConvolutionalLayer {cMap, (5:5), pad=true, activation=ReLU}(inputFeatures)
            pool1   = MaxPoolingLayer {(4:4), stride=(4:4)}(conv1)
            unpool1 = MaxUnpoolingLayer {(4:4), stride=(4:4)}(pool1, conv1)
            deconv1 = ConvolutionTransposeLayer {1, (5:5), cMap, pad=true, bias=false}(unpool1)
        }.deconv1

        # inputs
        imageShape = 28:28:1
        features = Input {imageShape}

        featScale = 1/256
        Scale{f} = x => Constant(f) .* x

        # apply model to features
        f1 = Scale{featScale} (features)
        z = model (f1)

        # rmse loss function
        f2 = Scale{featScale} (features)
        err = z - f2
        sqErr = ElementTimes(err, err)
        mse = ReduceMean(sqErr)
        rmse = Sqrt(mse)
        
        # declare special nodes
        featureNodes = (features)
        criterionNodes = (rmse)
        evaluationNodes = (rmse)
        outputNodes = (z)
    }

    SGD = {
        epochSize = 60000
        minibatchSize = 64
        maxEpochs = 3
        
        learningRatesPerSample = 0.00015
        momentumAsTimeConstant = 600
        
        firstMBsToShowResult = 5
        numMBsToShowResult = 235
    }

    reader = {
        readerType = "CNTKTextFormatReader"
        # See DataSets/MNIST/README.md for details on getting the data (Train-28x28_cntk_text.txt).
        file = "$DataDir$/Train-28x28_cntk_text.txt"
        input = {
            features = { dim = 784 ; format = "dense" }
            labels =   { dim = 10  ; format = "dense" }
        }
    }   
}

# TEST CONFIG
testNetwork = {
    action = "test"
    minibatchSize = 1024    # reduce this if you run out of memory

    reader = {
        readerType = "CNTKTextFormatReader"
        file = "$DataDir$/Test-28x28_cntk_text.txt"
        input = {
            features = { dim = 784 ; format = "dense" }
            labels =   { dim = 10  ; format = "dense" }
        }
    }
}

# WRITE CONFIG
writeResults = {
    action = "write"
    minibatchSize = 1
    outputPath = "$outputDir$/decoder_output_bs.txt"
    
    reader = {
        randomize = False
        readerType = "CNTKTextFormatReader"
        file = "$DataDir$/Test-28x28_cntk_text.txt"
        input = {
            features = { dim = 784 ; format = "dense" }
            labels =   { dim = 10  ; format = "dense" }
        }
    }
}