Neural networks and shifter romance
Jul. 13th, 2018 12:35 pmI found an online neural network (via following a link from aiweirdness.com) that I could feed stuff into in order to produce machine generated stuff.
So I grabbed the titles of 2000 shifter romance books from Goodreads lists and fed them in.
The first time I did it, I let it run for about 50 epochs (an epoch is one run through the data source), and it produced, well... let's say that it learned that capital letters should come at the beginnings of words and it had an inordinate fondness for dragons.
( First results )
My next try at it was a bit better. I had it run for 200 epochs and had it spit back results every 20.
( Read more... )
If you'd like to see my entire list of results, I've got them in a Google doc here: https://docs.google.com/document/d/1CSI9LLPaAAWHZDY9lXxYYaEtsU7KZ9tT5oIfvqC9Fzk/edit?usp=sharing
So I grabbed the titles of 2000 shifter romance books from Goodreads lists and fed them in.
The first time I did it, I let it run for about 50 epochs (an epoch is one run through the data source), and it produced, well... let's say that it learned that capital letters should come at the beginnings of words and it had an inordinate fondness for dragons.
( First results )
My next try at it was a bit better. I had it run for 200 epochs and had it spit back results every 20.
( Read more... )
If you'd like to see my entire list of results, I've got them in a Google doc here: https://docs.google.com/document/d/1CSI9LLPaAAWHZDY9lXxYYaEtsU7KZ9tT5oIfvqC9Fzk/edit?usp=sharing