def plotGeneratedImages(epoch,example=100,dim=(10,10),figsize=(10,10)):
noise = np.random.normal(0,1,size=(example,randomDim))
generatedImage = generator.predict(noise)
generatedImage = generatedImage.reshape(example,28,28)
plt.figure(figsize=figsize)
for i in range(example):
plt.subplot(dim[0],dim[1],i+1)
plt.imshow(generatedImage[i],interpolation='nearest',cmap='gray')
'''drop the x and y axis'''
plt.axis('off')
plt.tight_layout()
if not os.path.exists('generated_image'):
os.mkdir('generated_image')
plt.savefig('generated_image/wgan_generated_img_epoch_%d.png' % epoch)
python类normal()的实例源码
def plotGeneratedImages(epoch,example=100,dim=(10,10),figsize=(10,10)):
noise = np.random.normal(0,1,size=(example,randomDim))
generatedImage = generator.predict(noise)
generatedImage = generatedImage.reshape(example,28,28)
plt.figure(figsize=figsize)
for i in range(example):
plt.subplot(dim[0],dim[1],i+1)
plt.imshow(generatedImage[i],interpolation='nearest',cmap='gray')
'''drop the x and y axis'''
plt.axis('off')
plt.tight_layout()
if not os.path.exists('generated_image'):
os.mkdir('generated_image')
plt.savefig('generated_image/wgan_generated_img_epoch_%d.png' % epoch)
def get_q_network(weights_path):
model = Sequential()
model.add(Dense(1024, init=lambda shape, name: normal(shape, scale=0.01, name=name), input_shape=(25112,)))
model.add(Activation('relu'))
model.add(Dropout(0.2))
model.add(Dense(1024, init=lambda shape, name: normal(shape, scale=0.01, name=name)))
model.add(Activation('relu'))
model.add(Dropout(0.2))
model.add(Dense(6, init=lambda shape, name: normal(shape, scale=0.01, name=name)))
model.add(Activation('linear'))
adam = Adam(lr=1e-6)
model.compile(loss='mse', optimizer=adam)
if weights_path != "0":
model.load_weights(weights_path)
return model
def initNormal(shape,name=None):
return initializations.normal(shape,scale=0.2,name=name)
def test_normal(tensor_shape):
_runner(initializations.normal, tensor_shape, target_mean=0., target_std=0.05)
def weights_init(shape, name=None, dim_ordering=None):
return normal(shape, scale=0.01, name=name)
def conv2D_init(shape, dim_ordering='tf', name=None):
return initializations.normal(shape, scale=0.02, dim_ordering=dim_ordering, name=name)
def unitary_svd_init(shape, name=None):
assert shape[0]==shape[1]
Re=initializations.normal(shape,scale=1.0,name=name).get_value()
Im=initializations.normal(shape,scale=1.0,name=name).get_value()
X = Re+1j*Im
[U,S,V]=np.linalg.svd(X)
X = np.dot(U,V)
ReX = np.real(X)
ImX = np.imag(X)
Xaug = np.concatenate([ReX,ImX],axis=0)
return K.variable(Xaug,name=name)
def test_normal(tensor_shape):
_runner(initializations.normal, tensor_shape, target_mean=0., target_std=0.05)
def my_init(shape, name=None):
return initializations.normal(shape, scale=0.1, name=name)
# Best val_loss: 0.0205 - val_acc: 0.9978 (just tried only once)
# 30 minutes on Amazon EC2 g2.2xlarge (NVIDIA GRID K520)
def embedding_learning(train_data, user_dic, artist_dic, context_list, n_users, n_items):
# User embeddings
UC = np.random.normal(0.0, 0.01, (n_users, dim_num))
# Item embeddings
IC = np.random.normal(0.0, 0.01, (n_items, dim_num))
try:
for iteration in range(max_iters):
print 'loading...iteration: %d'%iteration
t = time.time()
for each_data in train_data:
u_i, i, w_i = each_data
w_i = w_i ** dis_coef
# print artist_dic[i]
for u_j in context_list[u_i]:
IC[artist_dic[i]] += learning_rate * ((1 - sigmoid(w_i)) * 2 * alpha * (UC[user_dic[u_i]] - UC[user_dic[u_j]]) - 2 * lamda * IC[artist_dic[i]])
UC[user_dic[u_i]] += learning_rate * ((1 - sigmoid(w_i)) * 2 * alpha * (IC[artist_dic[i]] - UC[user_dic[u_i]]) - 2 * lamda * UC[user_dic[u_i]])
UC[user_dic[u_j]] += learning_rate * ((1 - sigmoid(w_i)) * 2 * alpha * (IC[artist_dic[i]] - UC[user_dic[u_j]]) - 2 * lamda * UC[user_dic[u_j]])
# print IC[artist_dic[i]]
print 'Iter: %d elapsed: %fseconds'%(iteration, time.time() - t)
finally:
np.save(model_dir + 'Item_Emb', IC)
np.save(model_dir + 'User_Emb', UC)
np.savetxt(model_dir + 'Item_Emb.txt', IC)
np.savetxt(model_dir + 'User_Emb.txt', UC)
print 'Model saved...'
def init_normal(shape, name=None):
return initializations.normal(shape, scale=0.01, name=name)
def initNormal(shape,name=None):
return initializations.normal(shape,scale=0.2,name=name)
def my_init(shape, name=None):
return initializations.normal(shape, scale=1.2, name=name)
def test_normal(tensor_shape):
_runner(initializations.normal, tensor_shape, target_mean=0., target_std=0.05)
def unitary_svd_init(shape, name=None):
assert shape[0]==shape[1]
Re=initializations.normal(shape,scale=1.0,name=name).get_value()
Im=initializations.normal(shape,scale=1.0,name=name).get_value()
X = Re+1j*Im
[U,S,V]=np.linalg.svd(X)
X = np.dot(U,V)
ReX = np.real(X)
ImX = np.imag(X)
Xaug = np.concatenate([ReX,ImX],axis=0)
return K.variable(Xaug,name=name)
def create_classifier(body, data, n_classes, l2_reg=0.):
# Include last layers
top = BatchNormalization(mode=0, axis=channel_idx, name="bn7")(body)
top = Activation('relu', name="relu7")(top)
top = AtrousConvolution2D(512, 3, 3, 'he_normal', atrous_rate=(12, 12),
border_mode='same', name="conv6a",
W_regularizer=l2(l2_reg))(top)
top = Activation('relu', name="conv6a_relu")(top)
name = "hyperplane_num_cls_%d_branch_%d" % (n_classes, 12)
def my_init(shape, name=None, dim_ordering='th'):
return initializations.normal(shape, scale=0.01, name=name)
top = AtrousConvolution2D(n_classes, 3, 3, my_init,
atrous_rate=(12, 12), border_mode='same',
name=name, W_regularizer=l2(l2_reg))(top)
top = Deconvolution2D(n_classes, 16, 16, top._keras_shape, bilinear_init,
'linear', border_mode='valid', subsample=(8, 8),
bias=False, name="upscaling_"+str(n_classes),
W_regularizer=l2(l2_reg))(top)
top = CropLayer2D(data, name='score')(top)
top = NdSoftmax()(top)
return top
# Create model of basic segnet
def create_actor_network(self, state_size,action_dim):
print("Now we build the model")
S = Input(shape=[state_size])
h0 = Dense(HIDDEN1_UNITS, activation='relu')(S)
h1 = Dense(HIDDEN2_UNITS, activation='relu')(h0)
Steering = Dense(1,activation='tanh',init=lambda shape, name: normal(shape, scale=1e-4, name=name))(h1)
Acceleration = Dense(1,activation='sigmoid',init=lambda shape, name: normal(shape, scale=1e-4, name=name))(h1)
Brake = Dense(1,activation='sigmoid',init=lambda shape, name: normal(shape, scale=1e-4, name=name))(h1)
V = merge([Steering,Acceleration,Brake],mode='concat')
model = Model(input=S,output=V)
return model, model.trainable_weights, S
def train(epochs=1,batchsize=128):
batchCount = X_train.shape[0] / batchsize
print 'Epochs',epochs
print 'Bathc_size',batchsize
print 'Batches per epoch',batchCount
#range ande xrange the different is a list and a generator
for e in xrange(1,epochs+1):
print '-'*15 , 'Epoch %d' % e , '-'*15
for _ in tqdm(xrange(batchCount)):
#Get a random set of input noise and images
noise = np.random.normal(0,1,size=[batchsize,randomDim])
imageBatch = X_train[np.random.randint(0,X_train.shape[0],size=batchsize)]
#generate fake MNIST images
generatedImages = generator.predict(noise)
#Default is axis=0, equal to vstack is concate up and down
X = np.concatenate([imageBatch,generatedImages])
#Labels for generated and real data
yDis = np.ones(2*batchsize)
#one-sided label smoothing
yDis[:batchsize] = -1
#Train discriminator
discriminator.trainable = True
dloss = discriminator.train_on_batch(X,yDis)
#Train generator
noise = np.random.normal(0,1,size=[batchsize,randomDim])
yGen = np.ones(batchsize) * -1
discriminator.trainable = False
gloss = gan.train_on_batch(noise,yGen)
'''
d_weight = discriminator.get_weights()
d_weight = clip_weight(d_weight,-0.01,0.01)
discriminator.set_weights(d_weight)
'''
#Store loss of most recent batch from this epoch
Dloss.append(dloss)
Gloss.append(gloss)
if e == 1 or e % 5 == 0:
plotGeneratedImages(e)
saveModels(e)
plot_loss(e)
def train(epochs=1,batchsize=128):
batchCount = X_train.shape[0] / batchsize
print 'Epochs',epochs
print 'Bathc_size',batchsize
print 'Batches per epoch',batchCount
#range ande xrange the different is a list and a generator
for e in xrange(1,epochs+1):
print '-'*15 , 'Epoch %d' % e , '-'*15
for _ in tqdm(xrange(batchCount)):
#Get a random set of input noise and images
noise = np.random.normal(0,1,size=[batchsize,randomDim])
imageBatch = X_train[np.random.randint(0,X_train.shape[0],size=batchsize)]
#generate fake MNIST images
generatedImages = generator.predict(noise)
#Default is axis=0, equal to vstack is concate up and down
X = np.concatenate([imageBatch,generatedImages])
#Labels for generated and real data
yDis = np.ones(2*batchsize)
#one-sided label smoothing
yDis[:batchsize] = -1
#Train discriminator
discriminator.trainable = True
dloss = discriminator.train_on_batch(X,yDis)
#Train generator
noise = np.random.normal(0,1,size=[batchsize,randomDim])
yGen = np.ones(batchsize) * -1
discriminator.trainable = False
gloss = gan.train_on_batch(noise,yGen)
'''
d_weight = discriminator.get_weights()
d_weight = clip_weight(d_weight,-0.01,0.01)
discriminator.set_weights(d_weight)
'''
#Store loss of most recent batch from this epoch
Dloss.append(dloss)
Gloss.append(gloss)
if e == 1 or e % 5 == 0:
plotGeneratedImages(e)
saveModels(e)
plot_loss(e)
