def construct_network(self, hidden_size, tied = True):
l = len(hidden_size)
input_dim = self.in_out[0]
feed = self.x
for j in xrange(l):
dim = hidden_size[j]
neural_layer = HiddenLayer(rng=self.rng,input=feed,n_in=input_dim,n_out=dim,layer_name= 'Hidden' + str(j),activation=tensor.nnet.sigmoid)
dA_layer = dA(numpy_rng=self.rng,
theano_rng=self.theano_rng,
input=feed,
n_visible=input_dim,
n_hidden=dim,
W=neural_layer.W,
bhid=neural_layer.b)
self.dA_layers.append(dA_layer)
self.neural_layers.append(neural_layer)
self.params.extend(neural_layer.params)
input_dim = dim
feed = neural_layer.output
# Adding the final softmax layer
self.output_layer = HiddenLayer(rng=self.rng,input=feed,n_in=hidden_size[len(hidden_size) - 1],n_out=self.in_out[1],layer_name= 'Softmax', activation=tensor.nnet.softmax)
self.params.extend(self.output_layer.params)
def test_dA(datasets, learning_rate=0.1, training_epochs=15,
batch_size=20, output_folder='dA_plots'):
"""
This demo is tested on MNIST
:type learning_rate: float
:param learning_rate: learning rate used for training the DeNosing
AutoEncoder
:type training_epochs: int
:param training_epochs: number of epochs used for training
:type dataset: string
:param dataset: path to the picked dataset
"""
train_set_x, train_set_y = datasets[0]
# compute number of minibatches for training, validation and testing
n_train_batches = train_set_x.get_value(borrow=True).shape[0] // batch_size
# start-snippet-2
# allocate symbolic variables for the data
index = T.lscalar() # index to a [mini]batch
x = T.matrix('x') # the data is presented as rasterized images
# end-snippet-2
if not os.path.isdir(output_folder):
os.makedirs(output_folder)
os.chdir(output_folder)
####################################
# BUILDING THE MODEL NO CORRUPTION #
####################################
rng = numpy.random.RandomState(123)
theano_rng = RandomStreams(rng.randint(2 ** 30))
da = dA(
numpy_rng=rng,
theano_rng=theano_rng,
input=x,
n_visible=28 * 28,
n_hidden=500
)
cost, updates = da.get_cost_updates(
corruption_level=0.,
learning_rate=learning_rate
)
train_da = theano.function(
[index],
cost,
updates = updates,
givens={
x: train_set_x[index * batch_size: (index + 1) * batch_size]
}
)
start_time = timeit.default_timer()
############
# TRAINING #
############
# go through training epochs
for epoch in range(training_epochs):
# go through trainng set
c = []
for batch_index in range(n_train_batches):
c.append(train_da(batch_index))
print('Training epoch %d, cost ' % epoch, numpy.mean(c))
end_time = timeit.default_timer()
training_time = (end_time - start_time)
print('The no corruption code for file ' +
os.path.split(__file__)[1] +
' ran for %.2fm' % ((training_time) / 60.))
# start-snippet-3
#####################################
# BUILDING THE MODEL CORRUPTION 30% #
#####################################
rng = numpy.random.RandomState(123)
theano_rng = RandomStreams(rng.randint(2 ** 30))
da = dA(
numpy_rng=rng,
theano_rng=theano_rng,
input=x,
n_visible=28 * 28,
n_hidden=500
)
cost, updates = da.get_cost_updates(
corruption_level=0.3,
learning_rate=learning_rate
)
train_da = theano.function(
[index],
cost,
updates=updates,
givens={
x: train_set_x[index * batch_size: (index + 1) * batch_size]
}
)
start_time = timeit.default_timer()
############
# TRAINING #
############
# go through training epochs
for epoch in range(training_epochs):
# go through trainng set
c = []
for batch_index in range(n_train_batches):
c.append(train_da(batch_index))
print('Training epoch %d, cost ' % epoch, numpy.mean(c))
end_time = timeit.default_timer()
training_time = (end_time - start_time)
print ('The 30% corruption code for file ' +
os.path.split(__file__)[1] +
' ran for %.2fm' % (training_time / 60.))
# end-snippet-3
# end-snippet-4
os.chdir('../')
