youngsend opened this issue 5 years ago · 0 comments
Dear Authors, In the second and third Conv2D, input_shape=img_shape is provided. Since they are not the first layer in the Model, is input_shape necessary here?