what is the meaning of adding input_size(_) in every layer of convolution networks and how does it work? doesnot the next layer gets automatically input from previous layers?
What is the meaning of adding input_size(_) in every layer of convolution network and how does it work?
hey @abubakar_nsit ,
Yeah you are right the next layer automatically gets the input size from previous connected layer , but if you are using sequential api to build models , then in the first layer you add it is necessary to add input_size as it tells the first layer about what input it is going to receive.
and If you are using Function Model Api , then there is no such need as at first you declare a input layer that works the same.
Some programmers use input_size to just have understanding for their own , how to layer worked and reshaped the input etc. Else , it is always recommended to use model.summary() to do this work.
I hope this would have resolved your doubt.
Thank You and Happy Learning 
.