This repo is a project submission for the Udacity Deep Learning Nanodegree Program. The goal of the project is to train a Recurrent Neural Network (RNN) to generate a Simpsons TV script. (It was trained on a small dataset, so the results aren't very sensible)
See this notebook for the details.
moe_szyslak:(into phone) moe's tavern. where the elite meet to drink.
nigel_bakerbutcher: yeah!
homer_simpson:(distraught) ah, homer, you gotta sell it just barney work in the back of my cruiser.
homer_simpson: yes! the prayer i forgot to say has been answered.
homer_simpson:(puzzled) what the?
moe_szyslak: sorry. the greatest entertainer in the world won't turn him over up.
homer_simpson: anyway, it's just as well close the dump.
mayor_joe_quimby: i'm gonna drink you under the table.
quimby_#2: no, i am going to do.
moe_szyslak:(on a duff luck) but a good one!
kent_brockman: and the worst name that's right! even a deliberate mis-statement of fact.(smiling) but there's no one could pull!
all:(ad lib) what's a great thing.
moe_szyslak: okay it's over!
homer_simpson:(perking up) wait a minute, all right.(large sob)
moe_szyslak: now, say hello to springfield's newest supermodel!
lenny_leonard: just be careful