python - “张量”对象没有属性 'lower'

Question

我正在用 14 个新类微调 MobileNet。当我通过以下方式添加新图层时:

x=mobile.layers[-6].output
x=Flatten(x)
predictions = Dense(14, activation='softmax')(x)
model = Model(inputs=mobile.input, outputs=predictions)

我得到错误:

'Tensor' object has no attribute 'lower'

同时使用:

model.compile(Adam(lr=.0001), loss='categorical_crossentropy', metrics=['accuracy'])
model.fit_generator(train_batches, steps_per_epoch=18,
                validation_data=valid_batches, validation_steps=3, epochs=60, verbose=2)

我得到错误:

Error when checking target: expected dense_1 to have 4 dimensions, but got array with shape (10, 14)

降低 是什么意思？我看到了其他微调脚本，除了在本例中为 x 的模型名称外，没有其他参数。

Answer 1

张量必须在调用层时传递给层，而不是作为参数。因此它必须是这样的:

x = Flatten()(x)  # first the layer is constructed and then it is called on x

为了更清楚，相当于这样:

flatten_layer = Flatten()  # instantiate the layer
x = flatten_layer(x)       # call it on the given tensor