为 Seq2Seq 模型添加注意力层

Question

我已经构建了编码器-解码器的 Seq2Seq 模型。我想为其添加一个注意力层。我尝试通过此添加注意力层，但没有帮助。

这是我最初的代码，没有注意

# Encoder
encoder_inputs = Input(shape=(None,))
enc_emb =  Embedding(num_encoder_tokens, latent_dim, mask_zero = True)(encoder_inputs)
encoder_lstm = LSTM(latent_dim, return_state=True)
encoder_outputs, state_h, state_c = encoder_lstm(enc_emb)
# We discard `encoder_outputs` and only keep the states.
encoder_states = [state_h, state_c]

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = Input(shape=(None,))
dec_emb_layer = Embedding(num_decoder_tokens, latent_dim, mask_zero = True)
dec_emb = dec_emb_layer(decoder_inputs)
# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
decoder_outputs, _, _ = decoder_lstm(dec_emb,
                                     initial_state=encoder_states)
decoder_dense = Dense(num_decoder_tokens, activation='softmax')
decoder_outputs = decoder_dense(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
model = Model([encoder_inputs, decoder_inputs], decoder_outputs)
model.summary()

这是我在解码器中添加注意层后的代码（编码器层与初始代码中的相同）

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = Input(shape=(None,))
dec_emb_layer = Embedding(num_decoder_tokens, latent_dim, mask_zero = True)
dec_emb = dec_emb_layer(decoder_inputs)
# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
attention = dot([decoder_lstm, encoder_lstm], axes=[2, 2])
attention = Activation('softmax')(attention)
context = dot([attention, encoder_lstm], axes=[2,1])
decoder_combined_context = concatenate([context, decoder_lstm])
decoder_outputs, _, _ = decoder_combined_context(dec_emb,
                                     initial_state=encoder_states)
decoder_dense = Dense(num_decoder_tokens, activation='softmax')
decoder_outputs = decoder_dense(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
model = Model([encoder_inputs, decoder_inputs], decoder_outputs)
model.summary()

执行此操作时，我收到错误

 Layer dot_1 was called with an input that isn't a symbolic tensor. Received type: <class 'keras.layers.recurrent.LSTM'>. Full input: [<keras.layers.recurrent.LSTM object at 0x7f8f77e2f3c8>, <keras.layers.recurrent.LSTM object at 0x7f8f770beb70>]. All inputs to the layer should be tensors.

有人可以帮忙在这个架构中安装一个注意力层吗？

Answer 1

点积需要在张量输出上计算...在编码器中您正确定义了编码器输出，在解码器中您必须添加decoder_outputs, state_h, state_c = decoder_lstm(enc_emb, initial_state=encoder_states)

现在的点积是

attention = dot([decoder_outputs, encoder_outputs], axes=[2, 2])
attention = Activation('softmax')(attention)
context = dot([attention, encoder_outputs], axes=[2,1])

连接不需要initial_states。你必须在你的 rnn 层中定义它：decoder_outputs, state_h, state_c = decoder_lstm(enc_emb, initial_state=encoder_states)

这是完整的例子

编码器+解码器

# dummy variables
num_encoder_tokens = 30
num_decoder_tokens = 10
latent_dim = 100

encoder_inputs = Input(shape=(None,))
enc_emb =  Embedding(num_encoder_tokens, latent_dim, mask_zero = True)(encoder_inputs)
encoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
encoder_outputs, state_h, state_c = encoder_lstm(enc_emb)
# We discard `encoder_outputs` and only keep the states.
encoder_states = [state_h, state_c]

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = Input(shape=(None,))
dec_emb_layer = Embedding(num_decoder_tokens, latent_dim, mask_zero = True)
dec_emb = dec_emb_layer(decoder_inputs)
# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
decoder_outputs, _, _ = decoder_lstm(dec_emb,
                                     initial_state=encoder_states)
decoder_dense = Dense(num_decoder_tokens, activation='softmax')
decoder_outputs = decoder_dense(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
model = Model([encoder_inputs, decoder_inputs], decoder_outputs)
model.summary()

带有注意力的解码器

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = Input(shape=(None,))
dec_emb_layer = Embedding(num_decoder_tokens, latent_dim, mask_zero = True)
dec_emb = dec_emb_layer(decoder_inputs)
# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = LSTM(latent_dim, return_sequences=True, return_state=True)
decoder_outputs, state_h, state_c = decoder_lstm(dec_emb, initial_state=encoder_states)
attention = dot([decoder_outputs, encoder_outputs], axes=[2, 2])
attention = Activation('softmax')(attention)
context = dot([attention, encoder_outputs], axes=[2,1])
decoder_outputs = concatenate([context, decoder_outputs])
decoder_dense = Dense(num_decoder_tokens, activation='softmax')(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
model = Model([encoder_inputs, decoder_inputs], decoder_dense)
model.summary()