def train_model(args):
"""Load the data, train the model, test the model, export / save the model
"""
torch.manual_seed(args.seed)
# Open our dataset
train_loader, test_loader = data_utils.load_data(args.test_split,
args.batch_size)
# Create the model
net = model.SonarDNN().double()
optimizer = optim.SGD(net.parameters(), lr=args.lr,
momentum=args.momentum, nesterov=False)
# Train / Test the model
for epoch in range(1, args.epochs + 1):
train(net, train_loader, optimizer, epoch)
test(net, test_loader)
# Export the trained model
torch.save(net.state_dict(), args.model_name)
if args.model_dir:
# Save the model to GCS
data_utils.save_model(args.model_dir, args.model_name)
def train_model(args):
"""Load the data, train the model, test the model, export / save the model
"""
torch.manual_seed(args.seed)
# Open our dataset
train_loader, test_loader = data_utils.load_data(args.test_split,
args.batch_size)
# Create the model
net = model.SonarDNN().double()
optimizer = optim.SGD(net.parameters(),
lr=args.lr,
momentum=args.momentum,
nesterov=False)
# Train / Test the model
for epoch in range(1, args.epochs + 1):
train(net, train_loader, optimizer, epoch)
test(net, test_loader)
# Export the trained model
torch.save(net.state_dict(), args.model_name)
if args.model_dir:
# Save the model to GCS
data_utils.save_model(args.model_dir, args.model_name)
def build():
print('Starting ...')
for i in range(epoch, 100):
print('Working on epoch ' + str(i+1)+' ...')
for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
train(trX[start:end], trY[start:end])
test_score = np.mean(np.argmax(teY, axis=1) == predict(teX))
if (i+1) % 1 == 0:
save_model(params, epoch=i+1, annotation=annotation, namestub='conv_net', test_score=test_score)
def train_model(args):
"""Load the data, train the model, test the model, export / save the model
"""
torch.manual_seed(args.seed)
# Open our dataset
train_loader, test_loader = data_utils.load_data(args.test_split,
args.seed,
args.batch_size)
# Create the model
net = model.SonarDNN().double()
optimizer = optim.SGD(net.parameters(),
lr=args.lr,
momentum=args.momentum,
nesterov=False)
# Train / Test the model
latest_accuracy = 0.0
for epoch in range(1, args.epochs + 1):
train(net, train_loader, optimizer)
latest_accuracy = test(net, test_loader)
# The default name of the metric is training/hptuning/metric.
# We recommend that you assign a custom name. The only functional
# difference is that if you use a custom name, you must set the
# hyperparameterMetricTag value in the HyperparameterSpec object in your
# job request to match your chosen name.
# https://cloud.google.com/ml-engine/reference/rest/v1/projects.jobs#HyperparameterSpec
hpt = hypertune.HyperTune()
hpt.report_hyperparameter_tuning_metric(
hyperparameter_metric_tag='my_accuracy_tag',
metric_value=latest_accuracy,
global_step=args.epochs)
# Export the trained model
torch.save(net.state_dict(), args.model_name)
if args.job_dir:
# Save the model to GCS
data_utils.save_model(args.job_dir, args.model_name)
else:
print('Accuracy: {:.0f}%'.format(latest_accuracy))
def train_model(args):
train_features, test_features, train_labels, test_labels = \
data_utils.load_data(args)
sonar_model = model.sonar_model()
sonar_model.fit(train_features, train_labels, epochs=args.epochs,
batch_size=args.batch_size)
score = sonar_model.evaluate(test_features, test_labels,
batch_size=args.batch_size)
print(score)
# Export the trained model
sonar_model.save(args.model_name)
if args.model_dir:
# Save the model to GCS
data_utils.save_model(args.model_dir, args.model_name)
def train_model(args):
"""Load the data, train the model, test the model, export / save the model
"""
torch.manual_seed(args.seed)
# Open our dataset
train_loader, test_loader = data_utils.load_data(
args.test_split, args.seed, args.batch_size)
# Create the model
net = model.SonarDNN().double()
optimizer = optim.SGD(net.parameters(), lr=args.lr,
momentum=args.momentum, nesterov=False)
# Train / Test the model
latest_accuracy = 0.0
for epoch in range(1, args.epochs + 1):
train(net, train_loader, optimizer)
latest_accuracy = test(net, test_loader)
# The default name of the metric is training/hptuning/metric.
# We recommend that you assign a custom name. The only functional
# difference is that if you use a custom name, you must set the
# hyperparameterMetricTag value in the HyperparameterSpec object in your
# job request to match your chosen name.
# https://cloud.google.com/ml-engine/reference/rest/v1/projects.jobs#HyperparameterSpec
hpt = hypertune.HyperTune()
hpt.report_hyperparameter_tuning_metric(
hyperparameter_metric_tag='my_accuracy_tag',
metric_value=latest_accuracy,
global_step=args.epochs)
# Export the trained model
torch.save(net.state_dict(), args.model_name)
if args.job_dir:
# Save the model to GCS
data_utils.save_model(args.job_dir, args.model_name)
else:
print('Accuracy: {:.0f}%'.format(latest_accuracy))
training_cutoff=training_agg_cutoff, validation_cutoff=training_cutoff)
_, _, df_train_ret, df_validation_ret = load_data(
training_cutoff=validation_cutoff, validation_cutoff=None)
#%% Train and Save Models
models = []
regions_val = list(set(df_validation[region_col]))
dates_val = sorted(list(set(df_validation[date_col])))
validation_predictions = {}
if train_bilstm:
bilstm = BILSTMModel(**bilstm_params_dict)
bilstm.fit(df_train, 'first')
models.append('bilstm')
save_model(bilstm, bilstm_file.replace(".pickle", "_train.pickle"))
if load_bilstm:
bilstm = load_model(bilstm_file.replace(".pickle", "_train.pickle"))
validation_predictions['bilstm'] = bilstm.predict(
regions_val, dates_val, 'first')
models.append('bilstm')
if train_sir:
sir = SIRModel(**sir_params_dict)
sir.fit(df_train)
save_model(sir, sir_file.replace(".pickle", "_train.pickle"))
if load_sir:
sir = load_model(sir_file.replace(".pickle", "_train.pickle"))
try:
validation_predictions['sir'] = sir.predict(regions_val, dates_val)
# Build Tokenizer and turn training documents into integer tokens
tok = Tokenizer(num_tokens=None, stop_words=stop_words_custom)
tokenized_train = tok.fit_transform(training_documents)
# Convert training samples and labels to numpy arrays
X = list_to_numpy(tokenized_train, tok)
y = np.asarray(training_labels)
# Split off developmental data
# Fixed random_state = 42 for DEBUG purposes
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.25,
random_state=49)
# Fit model on training data
nb_clf = MultinomialNB()
nb_clf.fit(X_train, y_train, alpha=0.9)
# Make predictions on developmental data
y_pred = nb_clf.predict(X_test)
# Print F1 score for each of the four classes, and overall accuracy
print(f1_score(y_test, y_pred, average=None))
print(accuracy_score(y_test, y_pred))
# Save model parameters --- priors and conditional probabilities
save_model("nbmodel.txt", nb_clf, tok)
# Save Tokenizer; will need it for test data
with open("tok.pickle", "wb") as g:
dill.dump(tok, g)
def train_flair(model_name=''):
# Init data
train_dataset, val_dataset = prepare_datasets_FLAIR()
train_loader = DataLoader(train_dataset, batch_size=10, shuffle=True)
val_loader = DataLoader(val_dataset, batch_size=10, shuffle=True)
loaders = dict(train=train_loader, val=val_loader)
# Init Model
if model_name == '':
model = UNetFLAIR().cuda()
optimizer = torch.optim.Adam(model.parameters(), lr=1e-3, amsgrad=True)
scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer=optimizer,
gamma=0.984)
loss_fn = nn.BCELoss()
else:
model = data_utils.load_model(model_name)
epochs = 500
epoch_losses = dict(train=[], val=[])
for epoch in range(epochs):
for phase in 'train val'.split():
if phase == 'train':
model = model.train()
torch.set_grad_enabled(True)
else:
model = model.eval()
torch.set_grad_enabled(False)
loader = loaders[phase]
running_loss = []
for batch in loader:
imgs, masks = batch
imgs = imgs.cuda()
masks = masks.cuda()
outputs = model(imgs)
loss = loss_fn(outputs, masks)
running_loss.append(loss.item())
if phase == 'train':
optimizer.zero_grad()
loss.backward()
optimizer.step()
# End of Epoch
print(f'{epoch}) {phase} loss: {np.mean(running_loss)}')
visualize_results(loader, model, epoch, phase)
if epoch % 10 == 0:
results_dir = 'weight_flair/'
if not os.path.isdir(results_dir):
os.makedirs(results_dir)
data_utils.save_model(model, results_dir + f'model_{epoch}.pt')
epoch_losses[phase].append(np.mean(running_loss))
if phase == 'val':
df = pd.DataFrame(data=epoch_losses)
df.to_csv('loss.csv')
tensorboard(epoch_losses[phase], phase)
if phase == 'train':
scheduler.step()
df, df_train, df_validation = load_data(
validation_cutoff=validation_cutoff)
_, df_train_agg, df_validation_agg = load_data(
training_cutoff=training_agg_cutoff, validation_cutoff=training_cutoff)
# %% Train and Save Models
models = []
regions_val = list(set(df_validation[region_col]))
dates_val = list(set(df_validation[date_col]))
validation_predictions = {}
if train_sir:
sir = SIRModel(**sir_params_dict)
sir.fit(df_train)
save_model(sir, sir_file)
if load_sir:
sir = load_model(sir_file)
try:
validation_predictions['sir'] = sir.predict(regions_val, dates_val)
except:
pass
models.append('sir')
if train_knn:
knn = KNNModel(**knn_params_dict)
knn.fit(df_train)
models.append('knn')
save_model(knn, knn_file)
if load_knn:
def test_mlp(learning_rate=0.01, L1_reg=0.00, L2_reg=0.0001, n_epochs=n_epochs_mlp,
dataset='mnist.pkl.gz', batch_size=20, n_hidden=500, randomInit=randomInit,
logfilename=logfilename, loadparams = loadparams, paramsfilename=paramsfilename,
testrun = testrun, add_blurs = add_blurs):
"""
Demonstrate stochastic gradient descent optimization for a multilayer
perceptron
This is demonstrated on MNIST.
:type learning_rate: float
:param learning_rate: learning rate used (factor for the stochastic
gradient
:type L1_reg: float
:param L1_reg: L1-norm's weight when added to the cost (see
regularization)
:type L2_reg: float
:param L2_reg: L2-norm's weight when added to the cost (see
regularization)
:type n_epochs: int
:param n_epochs: maximal number of epochs to run the optimizer
:type dataset: string
:param dataset: the path of the MNIST dataset file from
http://www.iro.umontreal.ca/~lisa/deep/data/mnist/mnist.pkl.gz
"""
print ('loadparams is ' + str(loadparams))
datasets = load_data(dataset)
train_set_x, train_set_y = datasets[0]
valid_set_x, valid_set_y = datasets[1]
test_set_x, test_set_y = datasets[2]
# compute number of minibatches for training, validation and testing
n_train_batches = train_set_x.get_value(borrow=True).shape[0] // batch_size
n_valid_batches = valid_set_x.get_value(borrow=True).shape[0] // batch_size
n_test_batches = test_set_x.get_value(borrow=True).shape[0] // batch_size
######################
# BUILD ACTUAL MODEL #
######################
print('... building the model')
# 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
y = T.ivector('y') # the labels are presented as 1D vector of
# [int] labels
randgen = 1234
if randomInit:
randgen = 3421
rng = numpy.random.RandomState(randgen)
# construct the MLP class
classifier = MLP(
rng=rng,
input=x,
n_in=28 * 28,
n_hidden=n_hidden,
n_out=10,
randomInit=randomInit,
loadparams = loadparams,
paramsfilename = paramsfilename
)
# start-snippet-4
# the cost we minimize during training is the negative log likelihood of
# the model plus the regularization terms (L1 and L2); cost is expressed
# here symbolically
cost = (
classifier.negative_log_likelihood(y)
+ L1_reg * classifier.L1
+ L2_reg * classifier.L2_sqr
)
# end-snippet-4
# compiling a Theano function that computes the mistakes that are made
# by the model on a minibatch
test_model = theano.function(
inputs=[index],
outputs=classifier.errors(y),
givens={
x: test_set_x[index * batch_size:(index + 1) * batch_size],
y: test_set_y[index * batch_size:(index + 1) * batch_size]
}
)
validate_model = theano.function(
inputs=[index],
outputs=classifier.errors(y),
givens={
x: valid_set_x[index * batch_size:(index + 1) * batch_size],
y: valid_set_y[index * batch_size:(index + 1) * batch_size]
}
)
# start-snippet-5
# compute the gradient of cost with respect to theta (sorted in params)
# the resulting gradients will be stored in a list gparams
gparams = [T.grad(cost, param) for param in classifier.params]
# specify how to update the parameters of the model as a list of
# (variable, update expression) pairs
# given two lists of the same length, A = [a1, a2, a3, a4] and
# B = [b1, b2, b3, b4], zip generates a list C of same size, where each
# element is a pair formed from the two lists :
# C = [(a1, b1), (a2, b2), (a3, b3), (a4, b4)]
updates = [
(param, param - learning_rate * gparam)
for param, gparam in zip(classifier.params, gparams)
]
# compiling a Theano function `train_model` that returns the cost, but
# in the same time updates the parameter of the model based on the rules
# defined in `updates`
train_model = theano.function(
inputs=[index],
outputs=cost,
updates=updates,
givens={
x: train_set_x[index * batch_size: (index + 1) * batch_size],
y: train_set_y[index * batch_size: (index + 1) * batch_size]
}
)
# end-snippet-5
###############
# TRAIN MODEL #
###############
print('... training')
# early-stopping parameters
# CCC Commenting out patience for simplicity and transparency's sake
# patience = 10000 # look as this many examples regardless
# patience_increase = 2 # wait this much longer when a new best is
# # found
# improvement_threshold = 0.995 # a relative improvement of this much is
# considered significant
validation_frequency = n_train_batches # CCCmin(n_train_batches, patience // 2)
# go through this many
# minibatche before checking the network
# on the validation set; in this case we
# check every epoch
best_validation_loss = numpy.inf
best_iter = 0
test_score = 0.
start_time = timeit.default_timer()
epoch = 0
if loadparams:
epoch = epoch_from_filename(paramsfilename)
done_looping = False
while (epoch < n_epochs) and (not done_looping):
epoch = epoch + 1
for minibatch_index in range(n_train_batches):
minibatch_avg_cost = train_model(minibatch_index)
# iteration number
iter = (epoch - 1) * n_train_batches + minibatch_index
if (iter + 1) % validation_frequency == 0:
# compute zero-one loss on validation set
validation_losses = [validate_model(i) for i
in range(n_valid_batches)]
this_validation_loss = numpy.mean(validation_losses)
print(
'epoch %i, minibatch %i/%i, validation error %f %%' %
(
epoch,
minibatch_index + 1,
n_train_batches,
this_validation_loss * 100.
)
)
# if we got the best validation score until now
if this_validation_loss < best_validation_loss:
#improve patience if loss improvement is good enough
# CCC if (
# this_validation_loss < best_validation_loss *
# improvement_threshold
# ):
# patience = max(patience, iter *patience_increase)
best_validation_loss = this_validation_loss
best_iter = iter
# test it on the test set
test_losses = [test_model(i) for i
in range(n_test_batches)]
test_score = numpy.mean(test_losses)
print((' epoch %i, minibatch %i/%i, test error of '
'best model %f %%') %
(epoch, minibatch_index + 1, n_train_batches,
test_score * 100.))
# CCC if patience <= iter:
# done_looping = True
# break
if epoch in saveepochs_mlp:
# test it on the test set
epoch_test_losses = [test_model(i) for i
in range(n_test_batches)]
epoch_test_score = numpy.mean(epoch_test_losses)
print(('epoch %i, test error of '
'best model %f %%') %
(epoch, epoch_test_score * 100.))
save_model(classifier.params, epoch, best_validation_loss, epoch_test_score,
'../data/models/best_model_mlp_'
, randomInit, add_blurs, testrun, logfilename, endrun = (n_epochs==epoch))
end_time = timeit.default_timer()
print(('Optimization complete. Best validation score of %f %% '
'obtained at iteration %i, with test performance %f %%') %
(best_validation_loss * 100., best_iter + 1, test_score * 100.))
# test it on the test set
final_test_losses = [test_model(i) for i
in range(n_test_batches)]
final_test_score = numpy.mean(final_test_losses)
print(('The final test score is %f %% ') %
(final_test_score * 100.))
print(('The code for file ' +
os.path.split(__file__)[1] +
' ran for %.2fm' % ((end_time - start_time) / 60.)), file=sys.stderr)
def evaluate_lenet5(learning_rate=0.1, n_epochs=n_epochs_convmlp, dataset='mnist.pkl.gz', nkerns=[20, 50],
batch_size=500, thislogfilename = logfilename,
loadparams=loadparams, paramsfilename=paramsfilename,
randomInit=False, testrun=testrun, add_blurs=add_blurs, blur=blur, rot_angles = rotation_angles, annotation =''):
""" Demonstrates lenet on MNIST dataset
:type learning_rate: float
:param learning_rate: learning rate used (factor for the stochastic
gradient)
:type n_epochs: int
:param n_epochs: maximal number of epochs to run the optimizer
:type dataset: string
:param dataset: path to the dataset used for training /testing (MNIST here)
:type nkerns: list of ints
:param nkerns: number of kernels on each layer
"""
loadedparams = [None] * 8
if loadparams:
print("Loading params from " + paramsfilename + "...")
loadedparams = load_params(paramsfilename)
rng = numpy.random.RandomState(23455)
datasets = load_data(dataset, add_the_blurs=add_blurs, blur = blur, angles = rot_angles)
if len(rot_angles)>0:
annotation += '_angles_'
for ang in rot_angles:
annotation += str(ang)+'_'
train_set_x, train_set_y = datasets[0]
valid_set_x, valid_set_y = datasets[1]
test_set_x, test_set_y = datasets[2]
# compute number of minibatches for training, validation and testing
n_train_batches = train_set_x.get_value(borrow=True).shape[0]
n_valid_batches = valid_set_x.get_value(borrow=True).shape[0]
n_test_batches = test_set_x.get_value(borrow=True).shape[0]
n_train_batches //= batch_size
n_valid_batches //= batch_size
n_test_batches //= batch_size
# allocate symbolic variables for the data
index = T.lscalar() # index to a [mini]batch
# start-snippet-1
x = T.matrix('x') # the data is presented as rasterized images
y = T.ivector('y') # the labels are presented as 1D vector of
# [int] labels
######################
# BUILD ACTUAL MODEL #
######################
print('... building the model')
# Reshape matrix of rasterized images of shape (batch_size, 28 * 28)
# to a 4D tensor, compatible with our LeNetConvPoolLayer
# (28, 28) is the size of MNIST images.
layer0_input = x.reshape((batch_size, 1, 28, 28))
# Construct the first convolutional pooling layer:
# filtering reduces the image size to (28-5+1 , 28-5+1) = (24, 24)
# maxpooling reduces this further to (24/2, 24/2) = (12, 12)
# 4D output tensor is thus of shape (batch_size, nkerns[0], 12, 12)
layer0 = LeNetConvPoolLayer(
rng,
input=layer0_input,
image_shape=(batch_size, 1, 28, 28),
filter_shape=(nkerns[0], 1, 5, 5),
poolsize=(2, 2),
W = loadedparams[6],
b=loadedparams[7]
)
# Construct the second convolutional pooling layer
# filtering reduces the image size to (12-5+1, 12-5+1) = (8, 8)
# maxpooling reduces this further to (8/2, 8/2) = (4, 4)
# 4D output tensor is thus of shape (batch_size, nkerns[1], 4, 4)
layer1 = LeNetConvPoolLayer(
rng,
input=layer0.output,
image_shape=(batch_size, nkerns[0], 12, 12),
filter_shape=(nkerns[1], nkerns[0], 5, 5),
poolsize=(2, 2),
W = loadedparams[4],
b = loadedparams[5]
)
# the HiddenLayer being fully-connected, it operates on 2D matrices of
# shape (batch_size, num_pixels) (i.e matrix of rasterized images).
# This will generate a matrix of shape (batch_size, nkerns[1] * 4 * 4),
# or (500, 50 * 4 * 4) = (500, 800) with the default values.
layer2_input = layer1.output.flatten(2)
# construct a fully-connected sigmoidal layer
layer2 = HiddenLayer(
rng,
input=layer2_input,
n_in=nkerns[1] * 4 * 4,
n_out=500,
activation=activation_convmlp,
W=loadedparams[2],
b=loadedparams[3]
)
# classify the values of the fully-connected sigmoidal layer
layer3 = LogisticRegression(input=layer2.output,
n_in=500, n_out=10,
W=loadedparams[0],
b=loadedparams[1])
# the cost we minimize during training is the NLL of the model
cost = layer3.negative_log_likelihood(y)
# create a function to compute the mistakes that are made by the model
test_model = theano.function(
[index],
layer3.errors(y),
givens={
x: test_set_x[index * batch_size: (index + 1) * batch_size],
y: test_set_y[index * batch_size: (index + 1) * batch_size]
}
)
validate_model = theano.function(
[index],
layer3.errors(y),
givens={
x: valid_set_x[index * batch_size: (index + 1) * batch_size],
y: valid_set_y[index * batch_size: (index + 1) * batch_size]
}
)
# create a list of all model parameters to be fit by gradient descent
params = layer3.params + layer2.params + layer1.params + layer0.params
# create a list of gradients for all model parameters
grads = T.grad(cost, params)
# train_model is a function that updates the model parameters by
# SGD Since this model has many parameters, it would be tedious to
# manually create an update rule for each model parameter. We thus
# create the updates list by automatically looping over all
# (params[i], grads[i]) pairs.
updates = [
(param_i, param_i - learning_rate * grad_i)
for param_i, grad_i in zip(params, grads)
]
train_model = theano.function(
[index],
cost,
updates=updates,
givens={
x: train_set_x[index * batch_size: (index + 1) * batch_size],
y: train_set_y[index * batch_size: (index + 1) * batch_size]
}
)
# end-snippet-1
###############
# TRAIN MODEL #
###############
print('... training')
# early-stopping parameters
# CCC Commenting out patience for simplicity and transparency's sake
# patience = 10000 # look as this many examples regardless
# patience_increase = 2 # wait this much longer when a new best is
# # found
# improvement_threshold = 0.995 # a relative improvement of this much is
# # considered significant
validation_frequency = n_train_batches #min(n_train_batches, patience // 2)
# go through this many
# minibatche before checking the network
# on the validation set; in this case we
# check every epoch
best_validation_loss = numpy.inf
best_iter = 0
test_score = 0.
start_time = timeit.default_timer()
epoch = 0
if loadparams:
epoch = epoch_from_filename(paramsfilename)
done_looping = False
while (epoch < n_epochs) and (not done_looping):
epoch = epoch + 1
for minibatch_index in range(n_train_batches):
iter = (epoch - 1) * n_train_batches + minibatch_index
if iter % 100 == 0:
print('training @ iter = ', iter)
cost_ij = train_model(minibatch_index)
if (iter + 1) % validation_frequency == 0:
# compute zero-one loss on validation set
validation_losses = [validate_model(i) for i
in range(n_valid_batches)]
this_validation_loss = numpy.mean(validation_losses)
print('epoch %i, minibatch %i/%i, validation error %f %%' %
(epoch, minibatch_index + 1, n_train_batches,
this_validation_loss * 100.))
# if we got the best validation score until now
if this_validation_loss < best_validation_loss:
#improve patience if loss improvement is good enough
# CCC if this_validation_loss < best_validation_loss * \
# improvement_threshold:
# patience = max(patience, iter * patience_increase)
# save best validation score and iteration number
best_validation_loss = this_validation_loss
best_iter = iter
# test it on the test set
test_losses = [
test_model(i)
for i in range(n_test_batches)
]
test_score = numpy.mean(test_losses)
print((' epoch %i, minibatch %i/%i, test error of '
'best model %f %%') %
(epoch, minibatch_index + 1, n_train_batches,
test_score * 100.))
# CCC if patience <= iter:
# done_looping = True
# break
if epoch in saveepochs_convmlp:
# test it on the test set
epoch_test_losses = [test_model(i) for i
in range(n_test_batches)]
epoch_test_score = numpy.mean(epoch_test_losses)
print(('epoch %i, test error of '
'best model %f %%') %
(epoch, epoch_test_score * 100.))
save_model(params, epoch, best_validation_loss, epoch_test_score, '../data/models/best_model_convolutional_mlp_'
, randomInit, add_blurs, testrun, thislogfilename, endrun = (n_epochs == epoch), annotation = annotation)
end_time = timeit.default_timer()
print('Optimization complete.')
print('Best validation score of %f %% obtained at iteration %i, '
'with test performance %f %%' %
(best_validation_loss * 100., best_iter + 1, test_score * 100.))
print(('The code for file ' +
os.path.split(__file__)[1] +
' ran for %.2fm' % ((end_time - start_time) / 60.)), file=sys.stderr)
