What the part 2 is doing? AI to write like Shakespeare?

[hcchengithub]

I can easily understand part 1, which is to recognize MNIST handwirtten digits all the way up to 99.51% accuracy. I enjoy experimenting all the tips, learning rate, dropout, up to BN. But can't see what is this part 2 doing at all. I appreciate anyone who point it out a little.

[martin-gorner]

@hcchengithub this model is a "language model". It predicts probabilities for the next character with a distribution similar to what you would find in the Shakespeare corpus. In itself, it is just for fun. Or you could call it an academic interest. But more useful models can be extrapolated from language models, like translation models or models that generate image captions. I explore some of them in this video: https://youtu.be/pzOzmxCR37I