Support differing batch sizes at train and test time

[scttl]

Due to the pre-allocation of intermediate matrices (like pre-activations), the settings used for batch sizes at training time, will automatically be expected at test time. This creates problems if we expect these to differ (ex. no mini-batch training where the batch size is the number of records).

[scttl]

As batch_size is no longer stored as part of model parameters, we are successfully able to train a model with one batch size, save it, load it, then use it to generate predictions of a different batch_size.