Привет всем, я получаю сообщение об ошибке при проверке ввода: ожидалось, что conv2d_42_input имеет 4 измерения, но получил массив с формой (646, 128, 128)

Question

Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_42 (Conv2D)           (None, 32, 128, 128)      320       
_________________________________________________________________
activation_64 (Activation)   (None, 32, 128, 128)      0         
_________________________________________________________________
conv2d_43 (Conv2D)           (None, 32, 126, 126)      9248      
_________________________________________________________________
activation_65 (Activation)   (None, 32, 126, 126)      0         
_________________________________________________________________
max_pooling2d_28 (MaxPooling (None, 32, 63, 63)        0         
_________________________________________________________________
dropout_37 (Dropout)         (None, 32, 63, 63)        0         
_________________________________________________________________
conv2d_44 (Conv2D)           (None, 64, 61, 61)        18496     
_________________________________________________________________
activation_66 (Activation)   (None, 64, 61, 61)        0         
_________________________________________________________________
max_pooling2d_29 (MaxPooling (None, 64, 30, 30)        0         
_________________________________________________________________
dropout_38 (Dropout)         (None, 64, 30, 30)        0         
_________________________________________________________________
flatten_13 (Flatten)         (None, 57600)             0         
_________________________________________________________________
dense_25 (Dense)             (None, 64)                3686464   
_________________________________________________________________
activation_67 (Activation)   (None, 64)                0         
_________________________________________________________________
dropout_39 (Dropout)         (None, 64)                0         
_________________________________________________________________
dense_26 (Dense)             (None, 4)                 260       
_________________________________________________________________
activation_68 (Activation)   (None, 4)                 0         
=================================================================
Total params: 3,714,788
Trainable params: 3,714,788
Non-trainable params: 0
_________________________________________________________________
Out[102]: True

Привет всем, я получаю Ошибка при проверке ввода: ожидалось, что conv2d_42_input будет иметь 4 измерения, но получил массив с формой (646, 128, 128).