def build(self, input_shape=None):
self.input_spec = layers.InputSpec(shape=input_shape)
if hasattr(self.layer, 'built') and not self.layer.built:
self.layer.build(input_shape)
# initialise p
self.p_logit = self.add_variable(name='p_logit',
shape=(1,),
initializer=tf.keras.initializers.random_uniform(self.init_min, self.init_max),
dtype=tf.float32,
trainable=True)
self.p = tf.nn.sigmoid(self.p_logit[0])
tf.add_to_collection("LAYER_P", self.p)
# initialise regulariser / prior KL term
input_dim = int(np.prod(input_shape[1:]))
weight = self.layer.kernel
kernel_regularizer = self.weight_regularizer * tf.reduce_sum(tf.square(
weight)) / (1. - self.p)
dropout_regularizer = self.p * tf.log(self.p)
dropout_regularizer += (1. - self.p) * tf.log(1. - self.p)
dropout_regularizer *= self.dropout_regularizer * input_dim
regularizer = tf.reduce_sum(kernel_regularizer + dropout_regularizer)
# Add the regularisation loss to collection.
tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES,
regularizer)
def build(self, input_shape):
self.input_spec = [kl.InputSpec(shape=input_shape)]
shape = [1 for _ in input_shape]
for i in self.axis:
shape[i] = input_shape[i]
self.gamma = self.add_weight(shape=shape,
initializer=self.gamma_init,
regularizer=self.gamma_regularizer,
name='gamma')
self.beta = self.add_weight(shape=shape,
initializer=self.beta_init,
regularizer=self.beta_regularizer,
name='beta')
self.built = True
def build(self, input_shape):
super(ResSASABasicBlock, self).build(input_shape)
self.input_spec = layers.InputSpec(shape=input_shape)
def __init__(self, padding=(1, 1)):
self.padding = tuple(padding)
self.input_spec = [layers.InputSpec(ndim=4)]
super().__init__()
def __init__(self, padding: Tuple[int, int] = (1, 1)) -> None:
super().__init__()
self.padding = tuple(padding)
self.input_spec = [layers.InputSpec(ndim=4)]