class AttLayer(Layer):
def __init__(self, **kwargs):
self.init = initializations.get('normal')
#self.input_spec = [InputSpec(ndim=3)]
super(AttLayer, self).__init__(** kwargs)
def build(self, input_shape):
assert len(input_shape)==3
#self.W = self.init((input_shape[-1],1))
self.W = self.init((input_shape[-1],))
#self.input_spec = [InputSpec(shape=input_shape)]
self.trainable_weights = [self.W]
super(AttLayer, self).build(input_shape) # be sure you call this somewhere!
def call(self, x, mask=None):
eij = K.tanh(K.dot(x, self.W))
ai = K.exp(eij)
weights = ai/K.sum(ai, axis=1).dimshuffle(0,'x')
weighted_input = x*weights.dimshuffle(0,1,'x')
return weighted_input.sum(axis=1)
def get_output_shape_for(self, input_shape):
return (input_shape[0], input_shape[-1])
我有兴趣从班级而不是 self.W (层的权重)中获取注意力权重。有人可以告诉我我该怎么做吗?
这是我所做的:
MAX_SENT_LENGTH=40
当我尝试将模型创建为:
sentEncoder =Model(sentence_input,weighted_inp)
它抛出以下错误:
模型的输出张量必须是 Keras 张量。找到: Sum{axis= 1 , acc_dtype=float64}.0