def ae2(data):
"""This one uses an Autoencode layer. This layer is a part of the graph and does not need a specific traget"""
miniBatchSize = 1
ls = MS.GradientDescent(lr=0.1)
cost = MC.MeanSquaredError()
i = ML.Input(8, name='inp')
h = ML.Hidden(3,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
name="hid")
o = ML.Autoencode(
i.name,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
learningScenario=ls,
costObject=cost,
name="out")
ae = i > h > o
# ae.init()
# o.train.printGraph()
for e in xrange(2000):
for i in xrange(0, len(data), miniBatchSize):
ae.train(o, inp=data[i:i + miniBatchSize])
return ae, o
def ae1(data):
'''Using a regression layer. This layer needs an explicit target'''
miniBatchSize = 2
ls = MS.GradientDescent(lr=0.1)
cost = MC.MeanSquaredError()
i = ML.Input(8, name='inp')
h = ML.Hidden(3,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
name="hid")
o = ML.Regression(
8,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
learningScenario=ls,
costObject=cost,
name="out")
ae = i > h > o
for e in xrange(1000):
for i in xrange(0, len(data), miniBatchSize):
ae.train(o,
inp=data[i:i + miniBatchSize],
targets=data[i:i + miniBatchSize])
return ae, o
def _setOutputs(self):
from theano.tensor.nnet import conv
for layer in self.network.inConnections[self]:
if self.inputs is None:
self.inputs = layer.outputs
else:
self.inputs += layer.outputs
if self.filterHeight > self.inputHeight:
raise ValueError(
"Filter height for '%s' cannot be bigger than its input height: '%s' > '%s'"
% (self.name, self.filterHeight, self.inputHeight))
if self.filterWidth > self.inputWidth:
raise ValueError(
"Filter width for '%s' cannot be bigger than its input width: '%s' > '%s'"
% (self.name, self.filterWidth, self.inputWidth))
self.convolution = conv.conv2d(
input=self.inputs,
filters=self.W,
filter_shape=self.getParameterShape('W'))
self.pooled = self.pooler.apply(self)
self.nbFlatOutputs = self.nbChannels * self.height * self.width
if self.b is None:
MI.ZeroBias().apply(self)
self.b = self.b.dimshuffle('x', 0, 'x', 'x')
self.outputs = self.pooled + self.b
self.testOutputs = self.pooled + self.b
def __init__(self, WInitialization=MI.SmallUniformWeights(), bInitialization=MI.ZeroBias(), epsilon=1e-6) : Decorator_ABC.__init__(self) self.epsilon = epsilon self.WInitialization = WInitialization self.bInitialization = bInitialization self.W = None self.b = None self.paramShape = None
def __init__(self,
size,
layerTypes,
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
**kwargs):
super(WeightBias_ABC, self).__init__(size,
layerTypes=layerTypes,
initializations=initializations,
**kwargs)
self.testInputs = None
self.parameters = {"W": None, "b": None}
def test_ae_reg(self):
data = []
for i in xrange(8):
zeros = numpy.zeros(8)
zeros[i] = 1
data.append(zeros)
ls = MS.GradientDescent(lr=0.1)
cost = MC.MeanSquaredError()
i = ML.Input(8, name='inp')
h = ML.Hidden(
3,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
name="hid")
o = ML.Regression(
8,
activation=MA.ReLU(),
initializations=[MI.SmallUniformWeights(),
MI.ZeroBias()],
learningScenario=ls,
costObject=cost,
name="out")
ae = i > h > o
miniBatchSize = 1
for e in xrange(2000):
for i in xrange(0, len(data), miniBatchSize):
ae.train(o,
inp=data[i:i + miniBatchSize],
targets=data[i:i + miniBatchSize])
res = ae.propagate(o, inp=data)["outputs"]
for i in xrange(len(res)):
self.assertEqual(numpy.argmax(data[i]), numpy.argmax(res[i]))
def __init__(self,
WInitialization=MI.SmallUniformWeights(),
bInitialization=MI.ZeroBias(),
epsilon=1e-6,
onTrain=True,
onTest=True):
Decorator_ABC.__init__(self)
self.epsilon = epsilon
self.WInitialization = WInitialization
self.bInitialization = bInitialization
self.W = None
self.b = None
self.paramShape = None
self.onTrain = onTrain
self.onTest = onTest
self.hyperParameters.extend(["onTrain", "onTest", "epsilon"])
* automatically saves the model if the training halts because of an error or if the process is killed
* saves a log if the process dies unexpectedly
* training results and hyper parameters values are recorded to a file
* allows you to define custom stop criteria
* training info is printed at each epoch, including best scores and at which epoch they were achieved
"""
if __name__ == "__main__":
# Let's define the network
ls = MS.GradientDescent(lr=0.01)
cost = MC.NegativeLogLikelihood()
i = ML.Input(28 * 28, name='inp')
h = ML.Hidden(500, activation=MA.Tanh(), initializations=[MI.GlorotTanhInit(), MI.ZeroBias()], regularizations=[MR.L1(0), MR.L2(0.0001)], name="hid")
o = ML.SoftmaxClassifier(10, learningScenario=ls, costObject=cost, name="out", regularizations=[MR.L1(0), MR.L2(0.0001)])
mlp = i > h > o
mlp.saveDOT("mnist_mlp")
mlp.saveHTML("mnist_mlp")
# And then map sets to the inputs and outputs of our network
train_set, validation_set, test_set = load_mnist()
trainData = MDM.Series(images=train_set[0], numbers=train_set[1])
trainMaps = MDM.DatasetMapper("train", miniBatchSize=500)
trainMaps.mapInput(i, trainData.images)
trainMaps.mapOutput(o, trainData.numbers)
testData = MDM.Series(images=test_set[0], numbers=test_set[1])
def __init__(self, size, layerType, initializations = [MI.HeWeights(), MI.ZeroBias()], **kwargs) :
super(WeightBias_ABC, self).__init__(size, layerType=layerType, initializations=initializations, **kwargs)
self.W = None
self.b = None
* allows you to define custom stop criteria
* training info is printed at each epoch, including best scores and at which epoch they were achieved
"""
if __name__ == "__main__":
# Let's define the network
ls = MS.GradientDescent(lr=0.01)
cost = MC.NegativeLogLikelihood()
i = ML.Input(28 * 28, name='inp')
h = ML.Hidden(500,
activation=MA.Tanh(),
initializations=[MI.GlorotTanhInit(),
MI.ZeroBias()],
regularizations=[MR.L1(0), MR.L2(0.0001)],
name="hid")
o = ML.SoftmaxClassifier(10,
learningScenario=ls,
costObject=cost,
name="out",
regularizations=[MR.L1(0),
MR.L2(0.0001)])
mlp = i > h > o
mlp.saveDOT("mnist_mlp")
mlp.saveHTML("mnist_mlp")
# And then map sets to the inputs and outputs of our network
train_set, validation_set, test_set = load_mnist()