C2W4_Assignment

Hi,
I have such model

Model: “sequential_9”


Layer (type) Output Shape Param #

conv2d_18 (Conv2D) (None, 26, 26, 4) 112

max_pooling2d_18 (MaxPoolin (None, 13, 13, 4) 0
g2D)

conv2d_19 (Conv2D) (None, 11, 11, 8) 296

max_pooling2d_19 (MaxPoolin (None, 5, 5, 8) 0
g2D)

flatten_7 (Flatten) (None, 200) 0

dense_18 (Dense) (None, 512) 102912

dense_19 (Dense) (None, 26) 13338

=================================================================
Total params: 116,658
Trainable params: 116,658
Non-trainable params: 0

On Epoch 1/15
I have error InvalidArgumentError with validation_data=validation_generator
Node: ‘sequential_10/conv2d_20/Conv2D’
input depth must be evenly divisible by filter depth: 1 vs 3
[[{{node sequential_10/conv2d_20/Conv2D}}]] [Op:__inference_train_function_6408]

What is wrong with validation_generator?

Thanks.

Best regards, Taras

Can you please send me your notebook on PM
Thank you
Regards,
Mayank Ghogale

Hi,
The problem was in the input_shape.
Correct value: input_shape=(28, 28, 1 )

2 Likes