TensorFlow 2.0: Tensor input reshape error thrown when input data size not divisible by batch size

Question

I\'m working on a project which implements the use of a custom built keras NN. The input data is passed though an embedding layer which projects it into n dimensional space.