This document discusses methods for compressing n-gram language models. It describes several key techniques: quantizing word IDs from 4 bytes to 2 bytes using Lloyd's algorithm; compressing trigram tables using Golomb coding; and representing n-grams succinctly using variable byte encoding for word IDs and succinct bit vectors. These methods can significantly reduce the size of n-gram language mod
![Compressed ngram](https://cdn-ak-scissors.b.st-hatena.com/image/square/aac12d0166efd0e63c5a81343f67c656c942aff0/height=288;version=1;width=512/https%3A%2F%2Fcdn.slidesharecdn.com%2Fss_thumbnails%2Fcompressedngram-110423010329-phpapp01-thumbnail.jpg%3Fwidth%3D640%26height%3D640%26fit%3Dbounds)