Tensorflow ImageNet weight、ValueError的自定义输入形状：输入必须具有静态方形形状（128、128、（160、160）、（192、192）或（224、224）之一）

产生错误的代码段：

base = MobileNet(input_shape=(150,150,3),
                                 include_top=False,
                                 weights='imagenet')

案例1：当我在tensorflow版本（

tf.\uuuuuu版本

）的PC系统中尝试上述代码时，=“

1.14.0

” 我没有收到任何错误，只是收到一个小小的警告说

C:\Users\Hp\anaconda3\lib\site-packages\keras_applications\mobilenet.py:207: UserWarning: `input_shape` is undefined or non-square, or `rows` is not in [128, 160, 192, 224]. 
Weights for input shape (224, 224) will be loaded as the default.
  warnings.warn('`input_shape` is undefined or non-square, '

但它是有效的，证明：-

base.summary（）

提供输出：-

_________________________________________________________________
Layer (type)                 Output Shape              Param #
=================================================================
input_1 (InputLayer)         [(None, 150, 150, 3)]     0
_________________________________________________________________
conv1_pad (ZeroPadding2D)    (None, 151, 151, 3)       0
_________________________________________________________________
conv1 (Conv2D)               (None, 75, 75, 32)        864
_________________________________________________________________
conv1_bn (BatchNormalization (None, 75, 75, 32)        128
_________________________________________________________________
conv1_relu (ReLU)            (None, 75, 75, 32)        0
_________________________________________________________________
conv_dw_1 (DepthwiseConv2D)  (None, 75, 75, 32)        288
_________________________________________________________________
conv_dw_1_bn (BatchNormaliza (None, 75, 75, 32)        128
_________________________________________________________________
conv_dw_1_relu (ReLU)        (None, 75, 75, 32)        0
_________________________________________________________________
conv_pw_1 (Conv2D)           (None, 75, 75, 64)        2048
_________________________________________________________________
conv_pw_1_bn (BatchNormaliza (None, 75, 75, 64)        256
_________________________________________________________________
conv_pw_1_relu (ReLU)        (None, 75, 75, 64)        0
_________________________________________________________________
......and so on....

一切都很完美

案例2：运行

tf的服务器机器出现问题。\uuuuu版本\uuuuu='1.12.0'

当我运行相同的代码段时，即

base = MobileNet(input_shape=(150,150,3),
                                 include_top=False,
                                 weights='imagenet')

它显示了上述错误，即：

ValueError: If imagenet weights are being loaded, input must have a static square shape (one of (128, 128), (160, 160), (192, 192), or (224, 224)). 
Input shape provided = (150, 150, 3)

如果这是1.12.0版本中如何修复的问题。 如果不起作用，为什么还要费心给include\u top=True/False？ 请解释一下