This is a ngram language model based on myself that utilises as few pre-made libraries as possible. This README can also guide you through setting up your own ngram model using your own discord package.
This repository contains an n-gram language model that I trained using my own discord messages. N-gram models work by predicting the next word based on an "n" number of words.
A unigram: 1 word at a time A bigram: 2 words at a time A trigram: 3 words at a time Etc, etc.
So it looks back at the previous word, and then predicts based on probability of the model what the next word would be.
I've derived a lot of my understanding off of the 3rd chapter of Speech and Language Processing by Daniel Jurafsky and James H. Martin
Daniel Jurafsky and James H. Martin. 2024. Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition with Language Models, 3rd edition. Online manuscript released August 20, 2024. https://web.stanford.edu/~jurafsky/slp3.
The model and messages have been removed but there is an implementation below: https://lily-bot-implementation.onrender.com/