ValueError：没有为任何变量提供渐变：['embedding/embeddings:0', ']答案

【问题标题】：ValueError: No gradients provided for any variable: ['embedding/embeddings:0', ']ValueError：没有为任何变量提供渐变：['embedding/embeddings:0', ']
【发布时间】：2020-10-24 03:15:37
【问题描述】：

我是 Tensorflow 2 的新手，我想在 keras/tensorflow 中训练一个多输入神经网络。这是我的示例代码：

First_inputs = Input(shape=(2000, ),name="first")
Second_inputs = Input(shape=(4, ),name="second")
embedding_layer = Embedding(3,3,  input_length=2000,)(First_inputs)
flatten = Flatten()(embedding_layer)
first_dense = Dense(neuronCount,kernel_initializer=initializer, )(flatten)
merge = concatenate([first_dense, Second_inputs])
drop = Dropout(dropout)(merge)
output = Dense(1, )(drop)
model = Model(inputs=[First_inputs, Second_inputs], outputs=output)
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.1,shuffle=True,  random_state=42)
First_inputs =x_train[:,0:2000]
Second_inputs =x_train[:,2000:2004]
model.fit(([First_inputs, Second_inputs], y_train),validation_data=([First_inputs, Second_inputs], y_train),verbose=1,epochs=100,steps_per_epoch=209)

但是，我收到此错误：

ValueError: No gradients provided for any variable: ['embedding/embeddings:0', 'dense/kernel:0', 'dense/bias:0', 'dense_1/kernel:0'].

有人知道问题出在哪里吗？谢谢！

【问题讨论】：

标签： python tensorflow keras deep-learning tensorflow2.0

【解决方案1】：

您的数据是 numpy 数组，您必须为 fit() 方法提供两个单独的参数，np.arrays 列表作为输入，np.array 作为标签。（删除元组作为输入）：

First_inputs = Input(shape=(2000, ),name="first")
Second_inputs = Input(shape=(4, ),name="second")
embedding_layer = Embedding(3,3,  input_length=2000,)(First_inputs)
flatten = Flatten()(embedding_layer)
first_dense = Dense(neuronCount,kernel_initializer=initializer, )(flatten)
merge = concatenate([first_dense, Second_inputs])
drop = Dropout(dropout)(merge)
output = Dense(1, )(drop)
model = Model(inputs=[First_inputs, Second_inputs], outputs=output)
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.1,shuffle=True,  
random_state=42)
First_inputs =x_train[:,0:2000]
Second_inputs =x_train[:,2000:2004]
model.fit([First_inputs, Second_inputs], y_train,validation_data=([First_inputs, 
Second_inputs], y_train),verbose=1,epochs=100,steps_per_epoch=209)

【讨论】：