def build_network(deepest=False):
dropout = [0., 0.1, 0.2, 0.3, 0.4]
conv = [(64, 3, 3), (128, 3, 3), (256, 3, 3), (512, 3, 3), (512, 2, 2)]
input = Input(shape=(32, 32, 3))
output = fractal_net(
c=3,
b=5,
conv=conv,
drop_path=0.15,
dropout=dropout,
#drop_path=0.15, dropout=None,
deepest=deepest)(input)
output = Flatten()(output)
#output = Dense(NB_CLASSES, init='he_normal')(output)
output = Dense(NB_CLASSES, kernel_initializer='he_normal')(output)
output = Activation('softmax')(output)
#model = Model(input=input, output=output)
model = Model(inputs=input, outputs=output)
optimizer = SGD(lr=LEARN_START, momentum=MOMENTUM)
#optimizer = RMSprop(lr=LEARN_START)
#optimizer = Adam()
#optimizer = Nadam()
model.compile(optimizer=optimizer,
loss='categorical_crossentropy',
metrics=['accuracy'])
#model.compile(optimizer=optimizer, loss='mean_squared_error', metrics=['accuracy'])
#plot(model, to_file='model.png')
return model
def build_network(deepest=False):
dropout = [0., 0.1, 0.2, 0.3, 0.4]
conv = [(64, 3, 3), (128, 3, 3), (256, 3, 3), (512, 3, 3), (512, 2, 2)]
input= Input(shape=(3, 32, 32) if K._BACKEND == 'theano' else (32, 32,3))
output = fractal_net(
c=3, b=5, conv=conv,
drop_path=0.15, dropout=dropout,
deepest=deepest)(input)
output = Flatten()(output)
output = Dense(NB_CLASSES, init='he_normal')(output)
output = Activation('softmax')(output)
model = Model(input=input, output=output)
#optimizer = SGD(lr=LEARN_START, momentum=MOMENTUM)
#optimizer = SGD(lr=LEARN_START, momentum=MOMENTUM, nesterov=True)
optimizer = Adam()
#optimizer = Nadam()
model.compile(optimizer=optimizer, loss='categorical_crossentropy', metrics=['accuracy'])
plot(model, to_file='model.png', show_shapes=True)
return model
def build_network(deepest = False):
dropout = [0., 0., 0., 0., 0.]
conv = [(64, 3, 3), (128, 3, 3), (256, 3, 3), (512, 3, 3), (512, 2, 2)]
input = Input(shape = (32, 32, 3))
output = fractal_net(c = 3, b = 5, conv = conv, drop_path = 0.15,
dropout = dropout, deepest = deepest)(input)
output = Flatten()(output)
output = Dense(512, kernel_initializer = 'he_normal')(output)
output = Activation('relu')(output)
output = Dropout(0.25)(output)
output = Dense(nb_classes, kernel_initializer = 'he_normal')(output)
output = Activation('softmax')(output)
model = Model(inputs = input, outputs = output)
optimizer = SGD(lr = lr, momentum = 0.9, nesterov = True)
# optimizer = Adam()
model.compile(optimizer = optimizer, loss = 'categorical_crossentropy', metrics= ['accuracy'])
# model.summary()
return model
def build_network(deepest=False):
dropout = [0., 0.1, 0.2, 0.3, 0.4]
conv = [(64, 3, 3), (128, 3, 3), (256, 3, 3), (512, 3, 3), (512, 2, 2)]
input = Input(shape=(3, 32, 32))
output = fractal_net(c=3,
b=5,
conv=conv,
drop_path=0.15,
dropout=dropout,
deepest=deepest)(input)
output = Flatten()(output)
output = Dense(NB_CLASSES, init='he_normal')(output)
output = Activation('softmax')(output)
model = Model(input=input, output=output)
#optimizer = SGD(lr=LEARN_START, momentum=MOMENTUM)
#optimizer = SGD(lr=LEARN_START, momentum=MOMENTUM, nesterov=True)
optimizer = Adam()
#optimizer = Nadam()
model.compile(optimizer=optimizer,
loss='categorical_crossentropy',
metrics=['accuracy'])
plot(model, to_file='model.png', show_shapes=True)
return model
def f(input):
return fractal_net(b=b,
c=c,
conv=b * [(nb_filter, conv_size)],
drop_path=drop_path,
dropout=b * [dropout])(input)