Kenneth Heafield

kenlm. KenLM: Faster and Smaller Language Model Queries

preprocess. Corpus preprocessing

intgemm. int8_t and int16_t matrix multiply based on https://arxiv.org/abs/1705.01991

fasterText. Library for fast text representation and classification.

nplm. Fork of http://nlg.isi.edu/software/nplm/ for threadsafety and efficiency.

MEMT. System Combination

lazy. A lazy decoder for syntax

mtplz. Code for the paper Faster Phrase-Based Decoding by Refining Feature State

runforest. A python library that manages and runs experimental pipelines.

SALM. Joy Zhang's suffix array language model

usage. Print resource usage of processes to stderr with LD_PRELOAD

elrc-scrape. Scrape ELRC-SHARE for corpora

moses-interp. Poorly implemented interpolation. Don't use this.

StreamException. C++ exceptions with streams

kenlm_regtest. Regression test data

azurehacks. Hacks to control Azure instances