OUR N-GRAM LANGUAGE MODEL

WAS BUILT IN PYTHON AND TRAINED ON LYRICS FROM BOWIE’S STUDIO ALBUMS FROM THE 70s

An N-gram model takes N number of words from a dataset and determines which words usually follow that sequence. When presented with a new word, it samples from that word’s spread of following words to pick what comes next. It then ‘slides’ its window of n words along the now created phrase and assesses the last N words in that phrase, beginning the process anew.

Made by Imogen and Graham. TO VIEW THE SOURCE CODE, PLEASE VISIT OUR GITHUB REPO.