Inconsistency in explanation of Batch size in lesson 7 on human numbers

In lesson 7 in human number notebook, when Jeremy explained bs (batch size) and bptt (back propagation through time) parameters of RNN he gave them different meaning than it was for other NN models before. But when I running notebook code now, I see that basically they have the same meaning in contrast to what Jeremy told in the lecture.
Was code of these notebooks changed after the lecture?
Or am I confusing something?