我使用的是 Keras 1.0。我的问题与这个(如何在 Keras 中实现均值池化层)相同，但那里的答案对我来说似乎不够充分。

我想实现这个网络:

以下代码不起作用:

sequence = Input(shape=(max_sent_len,), dtype='int32')embedded = Embedding(vocab_size, word_embedding_size)(sequence)lstm = LSTM(hidden_state_size, activation='sigmoid', inner_activation='hard_sigmoid', return_sequences=True)(embedded)pool = AveragePooling1D()(lstm)output = Dense(1, activation='sigmoid')(pool)

如果我不设置 return_sequences=True，当我调用 AveragePooling1D() 时会得到这个错误:

Traceback (most recent call last):  File "<stdin>", line 1, in <module>  File "/PATH/keras/engine/topology.py", line 462, in __call__    self.assert_input_compatibility(x)  File "/PATH/keras/engine/topology.py", line 382, in assert_input_compatibility    str(K.ndim(x)))Exception: ('Input 0 is incompatible with layer averagepooling1d_6: expected ndim=3', ' found ndim=2')

否则，当我调用 Dense() 时会得到这个错误:

Traceback (most recent call last):  File "<stdin>", line 1, in <module>  File "/PATH/keras/engine/topology.py", line 456, in __call__    self.build(input_shapes[0])  File "/fs/clip-arqat/mossaab/trec/liveqa/cmu/venv/lib/python2.7/site-packages/keras/layers/core.py", line 512, in build    assert len(input_shape) == 2AssertionError

回答：

添加 TimeDistributed(Dense(1)) 起到了帮助:

sequence = Input(shape=(max_sent_len,), dtype='int32')embedded = Embedding(vocab_size, word_embedding_size)(sequence)lstm = LSTM(hidden_state_size, activation='sigmoid', inner_activation='hard_sigmoid', return_sequences=True)(embedded)distributed = TimeDistributed(Dense(1))(lstm)pool = AveragePooling1D()(distributed)output = Dense(1, activation='sigmoid')(pool)