keras argmax 没有渐变。如何定义 argmax 的梯度？

Question

我在伽马层中使用Keras.Backend.armax()。模型编译正常，但在 fit() 期间抛出错误。

ValueError: An operation has `None` for gradient. Please make sure that all of your ops have a gradient defined (i.e. are differentiable). Common ops without gradient: K.argmax, K.round, K.eval.

我的模特：

latent_dim = 512
encoder_inputs = Input(shape=(train_data.shape[1],))
encoder_dense = Dense(vocabulary, activation='softmax')
encoder_outputs = Embedding(vocabulary, latent_dim)(encoder_inputs)
encoder_outputs = LSTM(latent_dim, return_sequences=True)(encoder_outputs)
encoder_outputs = Dropout(0.5)(encoder_outputs)
encoder_outputs = encoder_dense(encoder_outputs)
encoder_outputs = Lambda(K.argmax, arguments={'axis':-1})(encoder_outputs)
encoder_outputs = Lambda(K.cast, arguments={'dtype':'float32'})(encoder_outputs)

encoder_dense1 = Dense(train_label.shape[1], activation='softmax')
decoder_embedding = Embedding(vocabulary, latent_dim)
decoder_lstm1 = LSTM(latent_dim, return_sequences=True)
decoder_lstm2 = LSTM(latent_dim, return_sequences=True)
decoder_dense2 = Dense(vocabulary, activation='softmax')

decoder_outputs = encoder_dense1(encoder_outputs)
decoder_outputs = decoder_embedding(decoder_outputs)
decoder_outputs = decoder_lstm1(decoder_outputs)
decoder_outputs = decoder_lstm2(decoder_outputs)
decoder_outputs = Dropout(0.5)(decoder_outputs)
decoder_outputs = decoder_dense2(decoder_outputs)
model = Model(encoder_inputs, decoder_outputs)
model.summary()

便于可视化的模型摘要：

_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_7 (InputLayer)         (None, 32)                0         
_________________________________________________________________
embedding_13 (Embedding)     (None, 32, 512)           2018816   
_________________________________________________________________
lstm_19 (LSTM)               (None, 32, 512)           2099200   
_________________________________________________________________
dropout_10 (Dropout)         (None, 32, 512)           0         
_________________________________________________________________
dense_19 (Dense)             (None, 32, 3943)          2022759   
_________________________________________________________________
lambda_5 (Lambda)            (None, 32)                0         
_________________________________________________________________
lambda_6 (Lambda)            (None, 32)                0         
_________________________________________________________________
dense_20 (Dense)             (None, 501)               16533     
_________________________________________________________________
embedding_14 (Embedding)     (None, 501, 512)          2018816   
_________________________________________________________________
lstm_20 (LSTM)               (None, 501, 512)          2099200   
_________________________________________________________________
lstm_21 (LSTM)               (None, 501, 512)          2099200   
_________________________________________________________________
dropout_11 (Dropout)         (None, 501, 512)          0         
_________________________________________________________________
dense_21 (Dense)             (None, 501, 3943)         2022759   
=================================================================
Total params: 14,397,283
Trainable params: 14,397,283
Non-trainable params: 0
_________________________________________________________________

我在谷歌上搜索了解决方案，但几乎所有都是关于一个有缺陷的模型。有些人建议不要使用导致问题的函数。但是，如您所见，如果没有 K.argmax，我无法创建此模型（如果您知道任何其他方式，请告诉我）。我该如何解决这个问题，从而训练我的模型？

Answer 1

出于显而易见的原因，Argmax 函数没有梯度；那怎么定义呢？为了使您的模型正常工作，您需要使该层不可训练。根据this question（或the documentation），您需要将trainable = False 传递给您的图层。至于层权重（如果适用），您可能希望将其设置为单位矩阵。