Markov
Blog post: https://doulighan.wordpress.com/2017/05/29/markov-chainz/
Markov chain generator. Pretty self explanitory.
On the run file, you can specify the length of resulting chain, as well as setting the state to read in characters or words, as a boolean. You can switch the training file from 'tolkien' to 'proust'. Note that 'tolkien' is meant to generate names, so reading the file as 'word = false' will give the best results. The 'proust' file will have better results if read as 'word = true'.
Feel free to use your own training data. You can add a txt file and change the run configuration to read in your new text file.
All code is original. Have fun!