def test_convolution_layer(self):
"""Test a model with a single CNTK Convolution layer against the
equivalent ELL predictor. This verifies that the import functions
reshape and reorder values appropriately and that the equivalent ELL
layer produces comparable output
"""
# Create a Convolution CNTK layer with no bias or activation,
# auto-padding, stride of 1
convolutionLayer = Convolution((3, 3), 5, pad=(
True, True), strides=1, bias=False, init=0)
x = input((2, 3, 4)) # Input order for CNTK is channels, rows, columns
cntkModel = convolutionLayer(x)
# Create a test set of weights to use for both CNTK and ELL layers
# CNTK has these in filters, channels, rows, columns order
weightValues = np.arange(90, dtype=np.float_).reshape(5, 2, 3, 3)
# Set the weights
convolutionLayer.parameters[0].value = weightValues
# create an ELL Tensor from the cntk weights, which re-orders the
# weights and produces an appropriately dimensioned tensor
weightTensor = cntk_converters.\
get_float_tensor_from_cntk_convolutional_weight_parameter(
convolutionLayer.parameters[0])
# Create the equivalent ELL predictor
layerParameters = ell.LayerParameters(
# Input order for ELL is rows, columns, channels. Account for
# padding.
ell.TensorShape(3 + 2, 4 + 2, 2),
ell.ZeroPadding(1),
ell.TensorShape(3, 4, 5),
ell.NoPadding())
convolutionalParameters = ell.ConvolutionalParameters(3, 1, 0, 5)
layer = ell.FloatConvolutionalLayer(
layerParameters, convolutionalParameters, weightTensor)
predictor = ell.FloatNeuralNetworkPredictor([layer])
# Get the results for both
inputValues = np.arange(24, dtype=np.float32).reshape(2, 3, 4)
cntkResults = cntkModel(inputValues)
orderedCntkResults = cntk_converters.get_float_vector_from_cntk_array(
cntkResults)
orderedInputValues = cntk_converters.get_float_vector_from_cntk_array(
inputValues)
ellResults = predictor.Predict(orderedInputValues)
# Compare the results
np.testing.assert_array_equal(
orderedCntkResults, ellResults,
'results for Convolution layer do not match!')
# now run same over ELL compiled model
self.verify_compiled(
predictor, orderedInputValues, orderedCntkResults, "convolution",
"test")
def test_prelu_activation_layer(self):
"""Test a model with a single CNTK PReLU activation layer against the
equivalent ELL predictor. This verifies that the import functions
reshape and reorder values appropriately and that the equivalent ELL
layer produces comparable output
"""
# Create a test set of alpha parameters to use for both CNTK and ELL
# layers
# Input order for CNTK is channels, rows, columns
alphaValues = np.linspace(1, 2, num=16 * 10 * 10,
dtype=np.float32).reshape(16, 10, 10)
# create an ELL Tensor from the alpha parameters, which re-orders and
# produces an appropriately dimensioned tensor
alphaTensor = cntk_converters.\
get_float_tensor_from_cntk_convolutional_weight_value_shape(
alphaValues, alphaValues.shape)
inputValues = np.linspace(-5, 5, num=16 * 10 * 10,
dtype=np.float32).reshape(16, 10, 10)
# Evaluate a PReLU CNTK layer
x = input((16, 10, 10))
p = parameter(shape=x.shape, init=alphaValues, name="prelu")
cntkModel = param_relu(p, x)
# Create the equivalent ELL predictor
layerParameters = ell.neural.LayerParameters(
# Input order for ELL is rows, columns, channels
ell.math.TensorShape(10, 10, 16),
ell.neural.NoPadding(),
ell.math.TensorShape(10, 10, 16),
ell.neural.NoPadding())
layer = ell.neural.FloatPReLUActivationLayer(layerParameters,
alphaTensor)
predictor = ell.neural.FloatNeuralNetworkPredictor([layer])
cntkResults = cntkModel(inputValues)
orderedCntkResults = cntk_converters.get_float_vector_from_cntk_array(
cntkResults)
orderedInputValues = cntk_converters.get_float_vector_from_cntk_array(
inputValues)
ellResults = predictor.Predict(orderedInputValues)
# Compare the results
np.testing.assert_array_equal(
orderedCntkResults, ellResults,
'results for PReLU Activation layer do not match!')
# now run same over ELL compiled model
self.verify_compiled(predictor, orderedInputValues, orderedCntkResults,
"prelu_activation", "test")
def test_dense_layer(self):
"""Test a model with a single CNTK Dense layer against the equivalent
ELL predictor. This verifies that the import functions reshape and
reorder values appropriately and that the equivalent ELL layer
produces comparable output
"""
# Create a Dense CNTK layer with no bias or activation
denseLayer = Dense(5, bias=False)
x = input((2, 3, 4)) # Input order for CNTK is channels, rows, columns
cntkModel = denseLayer(x)
# Create a test set of weights to use for both CNTK and ELL layers
# CNTK has these in channels, rows, columns, [output shape] order
weightValues = np.arange(120, dtype=np.float_).reshape(2, 3, 4, 5)
# Set the weights
denseLayer.parameters[0].value = weightValues
# create an ELL Tensor from the cntk weights, which re-orders the
# weights and produces an appropriately dimensioned tensor
weightTensor = cntk_converters.\
get_float_tensor_from_cntk_dense_weight_parameter(
denseLayer.parameters[0])
# Create the equivalent ELL predictor
layerParameters = ell.neural.LayerParameters(
# Input order for ELL is rows, columns, channels
ell.math.TensorShape(3, 4, 2),
ell.neural.NoPadding(),
ell.math.TensorShape(1, 1, 5),
ell.neural.NoPadding())
layer = ell.neural.FloatFullyConnectedLayer(layerParameters,
weightTensor)
predictor = ell.neural.FloatNeuralNetworkPredictor([layer])
# Get the results for both
inputValues = np.arange(24, dtype=np.float32).reshape(2, 3, 4)
orderedInputValues = cntk_converters.get_float_vector_from_cntk_array(
inputValues)
cntkResults = cntkModel(inputValues)
orderedCntkResults = cntk_converters.get_float_vector_from_cntk_array(
cntkResults)
ellResults = predictor.Predict(orderedInputValues)
# Compare the results
np.testing.assert_array_equal(orderedCntkResults, ellResults,
'results for Dense layer do not match!')
# now run same over ELL compiled model
self.verify_compiled(predictor, orderedInputValues, orderedCntkResults,
"dense", "test")
def verify_ell(self, op_name, predictor, input_data, cntk_output):
"""Compare this with the equivalent ELL layer, both reference and
compiled.
`op_name` is a human-readable name describing the layer being verified.
"""
ellTestInput = cntk_converters.get_float_vector_from_cntk_array(
input_data)
ellResults = np.array(predictor.Predict(ellTestInput)).ravel()
ellResultsOutputShape = predictor.GetOutputShape()
reshapedEllOutput = np.reshape(
ellResults,
(ellResultsOutputShape.rows, ellResultsOutputShape.columns,
ellResultsOutputShape.channels))
if len(input_data.shape) == 3:
reshapedEllOutput = np.transpose(reshapedEllOutput,
(2, 0, 1)) # to match CNTK output
# now compare these results.
np.testing.assert_array_almost_equal(
cntk_output,
reshapedEllOutput.ravel(),
decimal=4,
err_msg=('Results for {} layer do not match!'.format(op_name)))
# and verify compiled is also the same
self.verify_compiled(predictor, ellTestInput, cntk_output, op_name)
def process(self, ellLayers):
"""Appends the ELL representation of the current layer to ellLayers."""
# Create the ell.LayerParameters for the ELL layer
layerParameters = ell.LayerParameters(
self.layer.ell_inputShape, self.layer.ell_inputPaddingParameters, self.layer.ell_outputShape, self.layer.ell_outputPaddingParameters)
bias = -1.0 * self.layer.constants[0].value
if len(bias.shape) == 0:
biasVector = converters.get_float_vector_from_constant(bias, layerParameters.outputShape.channels)
else:
biasVector = converters.get_float_vector_from_cntk_array(bias)
# Create the ELL bias layer
ellLayers.append(ell.FloatBiasLayer(layerParameters, biasVector))
def process(self, ellLayers):
"""Appends the ELL representation of the current layer to ellLayers."""
# Create the ELL.LayerParameters for the ELL layer
layerParameters = ELL.LayerParameters(
self.layer.ell_inputShape, self.layer.ell_inputPaddingParameters,
self.layer.ell_outputShape, self.layer.ell_outputPaddingParameters)
# Create ELL scaling layer
if (self.scale.value.size == 1):
scalesVector = converters.get_float_vector_from_constant(
self.scale.value, layerParameters.outputShape.channels)
else:
scalesVector = converters.get_float_vector_from_cntk_array(
self.scale.value)
ellLayers.append(ELL.FloatScalingLayer(layerParameters, scalesVector))
def save_layer_outputs(self, op_name):
name = op_name + "(" + str(self.layer_index) + ")"
self.layer_index += 1
self.report.write("## %s\n" % (name))
self.report.write("````\n")
shape = self.data.shape
self.report.write("Output size: " + str(shape))
self.report.write("````\n")
with open("Compare_" + name + ".csv", "w") as f:
f.write("cntk,ell,compiled\n")
a = cntk_converters.get_float_vector_from_cntk_array(self.data)
b = self.ell_data.ravel()
c = self.compiled_data.ravel()
pos = 0
while pos < len(a) and pos < len(b) and pos < len(c):
f.write("%f,%f,%f\n" % (a[pos], b[pos], b[pos]))
pos += 1
f.close()
def test_batch_normalization_layer(self):
"""Test a model with a single CNTK BatchNormalization layer against the
equivalent ELL predictor This verifies that the import functions
reshape and reorder values appropriately and that the equivalent ELL
layer produces comparable output
"""
# Create a test set of scales and biases to use for both CNTK and ELL
# layers
scaleValues = np.linspace(0.1, 0.5, num=16, dtype=np.float32)
scaleVector = cntk_converters.get_float_vector_from_cntk_array(
scaleValues)
biasValues = np.linspace(1, 2, num=16, dtype=np.float32)
biasVector = cntk_converters.get_float_vector_from_cntk_array(
biasValues)
meanValues = np.linspace(-0.5, 0.5, num=16, dtype=np.float32)
meanVector = cntk_converters.get_float_vector_from_cntk_array(
meanValues)
varianceValues = np.linspace(-1, 1, num=16, dtype=np.float32)
varianceVector = cntk_converters.get_float_vector_from_cntk_array(
varianceValues)
# Create a BatchNormalization CNTK layer
# CNTK's BatchNormalization layer does not support setting the running
# mean and variance, so we use a wrapper function around the
# batch_normalization op
batchNorm = BatchNormalizationTester(init_scale=scaleValues,
norm_shape=scaleValues.shape,
init_bias=biasValues,
init_mean=meanValues,
init_variance=varianceValues)
# Input order for CNTK is channels, rows, columns
x = input((16, 10, 10))
cntkModel = batchNorm(x)
# Create the equivalent ELL predictor
layers = []
layerParameters = ell.neural.LayerParameters(
# Input order for ELL is rows, columns, channels
ell.math.TensorShape(10, 10, 16),
ell.neural.NoPadding(),
ell.math.TensorShape(10, 10, 16),
ell.neural.NoPadding())
# CNTK BatchNorm = ELL's BatchNorm + Scaling + Bias
# 1e-5 is the default epsilon for CNTK's BatchNormalization Layer
epsilon = 1e-5
layers.append(
ell.neural.FloatBatchNormalizationLayer(
layerParameters, meanVector, varianceVector, epsilon,
ell.neural.EpsilonSummand.variance))
layers.append(
ell.neural.FloatScalingLayer(layerParameters, scaleVector))
layers.append(ell.neural.FloatBiasLayer(layerParameters, biasVector))
predictor = ell.neural.FloatNeuralNetworkPredictor(layers)
inputValues = np.linspace(-5, 5, num=16 * 10 * 10,
dtype=np.float32).reshape(16, 10, 10)
cntkResults = cntkModel(inputValues)
orderedCntkResults = cntk_converters.get_float_vector_from_cntk_array(
# Note that cntk inserts an extra dimension of 1 in the front
cntkResults)
orderedInputValues = cntk_converters.get_float_vector_from_cntk_array(
inputValues)
ellResults = predictor.Predict(orderedInputValues)
# Compare the results (precision is 1 less decimal place from epsilon)
np.testing.assert_array_almost_equal(
orderedCntkResults, ellResults, 6,
'results for BatchNormalization layer do not match!')
# now run same over ELL compiled model
self.verify_compiled(predictor,
orderedInputValues,
orderedCntkResults,
"batch_norm",
"test",
precision=6)
def test_binary_convolution_layer(self):
"""Test a model with a single CNTK Binary Convolution layer against the
equivalent ELL predictor. This verifies that the import functions
reshape and reorder values appropriately and that the equivalent ELL
layer produces comparable output
"""
# Create a test set of weights to use for both CNTK and ELL layers
# CNTK has these in filters, channels, rows, columns order
weightValues = np.random.uniform(low=-5, high=5,
size=(5, 2, 3, 3)).astype(np.float32)
# create an ELL Tensor from the cntk weights, which re-orders the
# weights and produces an appropriately dimensioned tensor
weightTensor = cntk_converters.\
get_float_tensor_from_cntk_convolutional_weight_value_shape(
weightValues, weightValues.shape)
# Create a Binary Convolution CNTK layer with no bias, no activation,
# stride 1
# Input order for CNTK is channels, rows, columns
x = input((2, 10, 10))
cntkModel = CustomSign(x)
cntkModel = BinaryConvolution((10, 10),
num_filters=5,
channels=2,
init=weightValues,
pad=True,
bias=False,
init_bias=0,
activation=False)(cntkModel)
# Create the equivalent ELL predictor
layerParameters = ell.neural.LayerParameters(
# Input order for ELL is rows, columns, channels. Account for
# padding.
ell.math.TensorShape(10 + 2, 10 + 2, 2),
ell.neural.ZeroPadding(1),
ell.math.TensorShape(10, 10, 5),
ell.neural.NoPadding())
convolutionalParameters = ell.neural.BinaryConvolutionalParameters(
3, 1, ell.neural.BinaryConvolutionMethod.bitwise,
ell.neural.BinaryWeightsScale.none)
layer = ell.neural.FloatBinaryConvolutionalLayer(
layerParameters, convolutionalParameters, weightTensor)
predictor = ell.neural.FloatNeuralNetworkPredictor([layer])
# Get the results for both
inputValues = np.random.uniform(low=-50, high=50,
size=(2, 10, 10)).astype(np.float32)
cntkResults = cntkModel(inputValues)
orderedCntkResults = cntk_converters.get_float_vector_from_cntk_array(
cntkResults)
orderedInputValues = cntk_converters.get_float_vector_from_cntk_array(
inputValues)
ellResults = predictor.Predict(orderedInputValues)
# Compare the results
np.testing.assert_array_equal(
orderedCntkResults, ellResults,
'results for Binary Convolution layer do not match!')
# now run same over ELL compiled model
self.verify_compiled(predictor, orderedInputValues, orderedCntkResults,
"binary_convolution", "test")
