本文整理汇总了Python中dragnn.python.network_units.add_var_initialized方法的典型用法代码示例。如果您正苦于以下问题：Python network_units.add_var_initialized方法的具体用法？Python network_units.add_var_initialized怎么用？Python network_units.add_var_initialized使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类dragnn.python.network_units的用法示例。
在下文中一共展示了network_units.add_var_initialized方法的3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: IdentityInitializerHelper
# 需要导入模块: from dragnn.python import network_units [as 别名]
# 或者: from dragnn.python.network_units import add_var_initialized [as 别名]
def IdentityInitializerHelper(self, shape, expected, divisor=1.0, std=1e-4):
"""Tests identity initialization by comparing expected to actual array.
Tests the given expected array against the result of calling
network_units.add_var_initialized() with the given params and
init_type='identity'.
Args:
shape: shape of the array
expected: expected contents of the array to initialize
divisor: numerator for identity initialization where the last two dims
of the array are not equal; should divide both of the last two dims
std: standard deviation for random normal samples
"""
with tf.Graph().as_default(), self.test_session() as session:
np.random.seed(4)
tensor = network_units.add_var_initialized('tensor', shape, 'identity',
divisor=divisor, stddev=std)
session.run(tf.global_variables_initializer())
actual = session.run(tensor)
self.assertAllClose(actual, expected, 1e-8, 1e-8) 示例2: __init__
# 需要导入模块: from dragnn.python import network_units [as 别名]
# 或者: from dragnn.python.network_units import add_var_initialized [as 别名]
def __init__(self, component):
super(PairwiseBilinearLabelNetwork, self).__init__(component)
parameters = component.spec.network_unit.parameters
self._num_labels = int(parameters['num_labels'])
self._source_dim = self._linked_feature_dims['sources']
self._target_dim = self._linked_feature_dims['targets']
self._weights = []
self._weights.append(
network_units.add_var_initialized('bilinear',
[self._source_dim,
self._num_labels,
self._target_dim],
'xavier'))
self._params.extend(self._weights)
self._regularized_weights.extend(self._weights)
self._layers.append(network_units.Layer(component,
name='bilinear_scores',
dim=self._num_labels)) 示例3: IdentityInitializerHelper
# 需要导入模块: from dragnn.python import network_units [as 别名]
# 或者: from dragnn.python.network_units import add_var_initialized [as 别名]
def IdentityInitializerHelper(self, shape, expected, divisor=1.0, std=1e-4):
"""Tests identity initialization by comparing expected to actual array.
Tests the given expected array against the result of calling
network_units.add_var_initialized() with the given params and
init_type='identity'.
Args:
shape: shape of the array
expected: expected contents of the array to initialize
divisor: numerator for identity initialization where the last two dims
of the array are not equal; should divide both of the last two dims
std: standard deviation for random normal samples
"""
with tf.Graph().as_default(), self.test_session() as session:
np.random.seed(4)
tensor = network_units.add_var_initialized(
'tensor', shape, 'identity', divisor=divisor, stddev=std)
session.run(tf.global_variables_initializer())
actual = session.run(tensor)
self.assertAllClose(actual, expected, 1e-8, 1e-8)