def __call__(self, w):
maximum_weight = K.max(K.abs(w))
w = w / (K.epsilon() + maximum_weight) # On [-1,1].
signs = K.sign(w)
unsigned_w = K.abs(w) # On [0,1].
edge_scaled_w_unsigned = unsigned_w * (
1. - self.min_value) + self.min_value # On [min_value,1].
edge_scaled_w = signs * edge_scaled_w_unsigned # On [-1,-min_value] U [min_value,1].
return edge_scaled_w
def __call__(self, w):
# First apply DivideByMax.
maximum_weight = K.max(K.abs(w))
w = w / (K.epsilon() + maximum_weight) # On [-1, 1].
# Then apply MinMaxNorm.
norms = K.sqrt(
math_ops.reduce_sum(math_ops.square(w),
axis=self.axis,
keepdims=True))
desired = (self.rate * K.clip(norms, self.min_value, self.max_value) +
(1 - self.rate) * norms)
return w * (desired / (K.epsilon() + norms))
def softmax(x, axis=-1):
"""Softmax activation function.
Arguments:
x : Tensor.
axis: Integer, axis along which the softmax normalization is applied.
Returns:
Tensor, output of softmax transformation.
Raises:
ValueError: In case `dim(x) == 1`.
"""
ndim = K.ndim(x)
if ndim == 2:
return K.softmax(x)
elif ndim > 2:
e = K.exp(x - K.max(x, axis=axis, keepdims=True))
s = K.sum(e, axis=axis, keepdims=True)
return e / s
else:
raise ValueError('Cannot apply softmax to a tensor that is 1D')
def sparse_top_k_categorical_accuracy(y_true, y_pred, k=5):
return K.mean(K.in_top_k(y_pred,
K.cast(K.max(y_true, axis=-1), 'int32'), k), axis=-1)
def sparse_categorical_accuracy(y_true, y_pred):
return K.cast(
K.equal(
K.max(y_true, axis=-1), K.cast(K.argmax(y_pred, axis=-1),
K.floatx())), K.floatx())
def call(self, inputs):
if self.data_format == 'channels_last':
return K.max(inputs, axis=[1, 2, 3])
else:
return K.max(inputs, axis=[2, 3, 4])
def call(self, inputs): return K.max(inputs, axis=1)
def call(self, inputs):
if self.data_format == 'channels_last':
return K.max(inputs, axis=[1, 2, 3])
else:
return K.max(inputs, axis=[2, 3, 4])
def call(self, inputs): return K.max(inputs, axis=1)
def categorical_hinge(y_true, y_pred): pos = K.sum(y_true * y_pred, axis=-1) neg = K.max((1. - y_true) * y_pred, axis=-1) return K.maximum(neg - pos + 1., 0.)
def call(self, inputs):
return backend.max(inputs, axis=1)
def sparse_top_k_categorical_accuracy(y_true, y_pred, k=5):
return K.mean(K.in_top_k(y_pred, K.cast(K.max(y_true, axis=-1), 'int32'),
k),
axis=-1)
def sparse_categorical_accuracy(y_true, y_pred):
return K.cast(
K.equal(K.max(y_true, axis=-1),
K.cast(K.argmax(y_pred, axis=-1), K.floatx())), K.floatx())
def call(self, inputs): return backend.max(inputs, axis=1)
def __call__(self, w):
maximum_weight = K.max(K.abs(w))
return w / (K.epsilon() + maximum_weight)
