curlywurlycraig / potterbot

Train a language model using subtitles from all the Harry Potter films.

Potterbot

Generate Harry Potter-like sentences. Trained using subtitles from http://www.yifysubtitles.com/search?q=harry+potter

Run using python potterbot.py

Installation

virtualenv .virtualenv --python=python3

.virtualenv/bin/pip install nltk

Sample Output

Potterbot: Dumbledore made you to the Restricted Section

Potterbot: There you will never forget this morning , Harry was cursed , Ron , I reckon you can be a kind of people 's a little cockroach

Potterbot: This is he 's a quiet to be sufficient quality of you can find it to sink no lower your time already been so what it well , you need any other arm , sir .

Potterbot: Please , go making the night , first task is a stopper in great ugly brute , but as much do you today , wait .

Potterbot: there then he says breakfast 's a troll in the blood will be learning how does n't know me .

Potterbot: - Dress robes ? But who would be counterproductive , obvious , Weasley is perfect I do n't worry .

Potterbot: He had to admit , yeah .

Potterbot: Furthermore , Petunia , enough for a secure the minute or all real loyalty down .

Potterbot: Even with the world will please join You-Know-Who is a broomstick , Ron 's Cross , we 'll go , everybody , Miss Granger .

Potterbot: Hideous

potterbot's Issues

README is unclear that nltk has an additional installation step

On initial installation, I had to follow steps here:
http://www.nltk.org/data.html

The README should include this in installation instructions, or, the app should self-install this if required. The latter is preferred.