本文整理汇总了Python中blocks.roles.WEIGHT属性的典型用法代码示例。如果您正苦于以下问题：Python roles.WEIGHT属性的具体用法？Python roles.WEIGHT怎么用？Python roles.WEIGHT使用的例子？那么, 这里精选的属性代码示例或许可以为您提供帮助。您也可以进一步了解该属性所在类blocks.roles的用法示例。
在下文中一共展示了roles.WEIGHT属性的7个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: _allocate
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def _allocate(self):
self.W_state = shared_floatx_nans((self.dim, 4*self.dim),
name='W_state')
self.W_cell_to_in = shared_floatx_nans((self.dim,),
name='W_cell_to_in')
self.W_cell_to_forget = shared_floatx_nans((self.dim,),
name='W_cell_to_forget')
self.W_cell_to_out = shared_floatx_nans((self.dim,),
name='W_cell_to_out')
# The underscore is required to prevent collision with
# the `initial_state` application method
self.initial_state_ = shared_floatx_zeros((self.dim,),
name="initial_state")
self.initial_cells = shared_floatx_zeros((self.dim,),
name="initial_cells")
add_role(self.W_state, WEIGHT)
add_role(self.W_cell_to_in, WEIGHT)
add_role(self.W_cell_to_forget, WEIGHT)
add_role(self.W_cell_to_out, WEIGHT)
add_role(self.initial_state_, INITIAL_STATE)
add_role(self.initial_cells, INITIAL_STATE)
self.parameters = [
self.W_state, self.W_cell_to_in, self.W_cell_to_forget,
self.W_cell_to_out, self.initial_state_, self.initial_cells] 示例2: _allocate
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def _allocate(self):
self.W_state = shared_floatx_nans((self.dim, 4.5 * self.dim),
name='W_state')
# The underscore is required to prevent collision with
# the `initial_state` application method
self.initial_state_ = shared_floatx_zeros((self.dim,),
name="initial_state")
self.initial_cells = shared_floatx_zeros((self.num_copies, self.dim),
name="initial_cells")
add_role(self.W_state, WEIGHT)
# add_role(self.initial_state_, INITIAL_STATE)
# add_role(self.initial_cells, INITIAL_STATE)
self.parameters = [self.W_state] 示例3: _allocate
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def _allocate(self):
W = shared_floatx_nans((self.input_dim, self.output_dim), name='W')
add_role(W, WEIGHT)
self.parameters.append(W)
self.add_auxiliary_variable(W.norm(2), name='W_norm')
if self.use_bias:
b = shared_floatx_nans((self.output_dim,), name='b')
add_role(b, BIAS)
self.parameters.append(b)
self.add_auxiliary_variable(b.norm(2), name='b_norm') 示例4: _allocate
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def _allocate(self):
self.parameters.append(shared_floatx_nans((self.length, self.dim),
name='W'))
add_role(self.parameters[-1], WEIGHT) 示例5: weight
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def weight(self, init, name, cast_float32=True, for_conv=False):
weight = self.shared(init, name, cast_float32, role=WEIGHT)
if for_conv:
return weight.dimshuffle('x', 0, 'x', 'x')
return weight 示例6: _allocate
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def _allocate(self):
self.parameters.append(shared_floatx_nans((self.dim, self.dim),
name='state_to_state'))
self.parameters.append(shared_floatx_nans((self.dim, 2 * self.dim),
name='state_to_gates'))
for i in range(2):
if self.parameters[i]:
add_role(self.parameters[i], WEIGHT) 示例7: main
# 需要导入模块: from blocks import roles [as 别名]
# 或者: from blocks.roles import WEIGHT [as 别名]
def main(save_to, num_epochs):
mlp = MLP([Tanh(), Softmax()], [784, 100, 10],
weights_init=IsotropicGaussian(0.01),
biases_init=Constant(0))
mlp.initialize()
x = tensor.matrix('features')
y = tensor.lmatrix('targets')
probs = mlp.apply(x)
cost = CategoricalCrossEntropy().apply(y.flatten(), probs)
error_rate = MisclassificationRate().apply(y.flatten(), probs)
cg = ComputationGraph([cost])
W1, W2 = VariableFilter(roles=[WEIGHT])(cg.variables)
cost = cost + .00005 * (W1 ** 2).sum() + .00005 * (W2 ** 2).sum()
cost.name = 'final_cost'
mnist_train = MNIST(("train",))
mnist_test = MNIST(("test",))
algorithm = GradientDescent(
cost=cost, parameters=cg.parameters,
step_rule=Scale(learning_rate=0.1))
extensions = [Timing(),
FinishAfter(after_n_epochs=num_epochs),
DataStreamMonitoring(
[cost, error_rate],
Flatten(
DataStream.default_stream(
mnist_test,
iteration_scheme=SequentialScheme(
mnist_test.num_examples, 500)),
which_sources=('features',)),
prefix="test"),
TrainingDataMonitoring(
[cost, error_rate,
aggregation.mean(algorithm.total_gradient_norm)],
prefix="train",
after_epoch=True),
Checkpoint(save_to),
Printing()]
if BLOCKS_EXTRAS_AVAILABLE:
extensions.append(Plot(
'MNIST example',
channels=[
['test_final_cost',
'test_misclassificationrate_apply_error_rate'],
['train_total_gradient_norm']]))
main_loop = MainLoop(
algorithm,
Flatten(
DataStream.default_stream(
mnist_train,
iteration_scheme=SequentialScheme(
mnist_train.num_examples, 50)),
which_sources=('features',)),
model=Model(cost),
extensions=extensions)
main_loop.run()