Hi all, I hope you are well,
I have a problem where I have a CNN with two inputs, one is a 75 x 4 matrix and the other one is a single value. With a batch size of 512, the size of the training batch matrix is naturally 512 x 75 x 4, and the network trains fine without the second input.
But, as soon as I try to add the vector of single values, each 1 x 75 x 4 Matrix is passed alongside the whole vector (512 entries) which raises an error.
During the training, I would like to forward pass each 1 x 75 x 4 matrix alongside its corresponding single value but I’ve yet to find how,
Thanks in advance,