14 Commits

Author	SHA1	Message	Date
MosesAdmin	f39252da9d	daily automatic beautifier	2016-09-05 00:00:38 +01:00
Paco Zamora Martinez	1c8b1d6056	Updated hashCode generation at NeuralLMWrapper.cpp; changed it to consider the last n-1 words instead of the whole ngram n words	2016-07-08 18:31:17 +02:00
michaelhutt	cf5d093c4f	fixes parameter mismatch in LM::Load	2015-12-23 13:31:49 -05:00
Ulrich Germann	fdb5d9b282	Code refactoring: All FF now get access to AllOptions at load time, so they don't have to rely on StaticData.	2015-12-08 12:56:55 +00:00
Nicola Bertoldi	e4eb201c52	merged master into dynamic-models and solved conflicts	2014-12-13 12:52:47 +01:00
Rico Sennrich	4a94a41310	compatibility with NPLM 0.3 (fork on https://github.com/rsennrich/nplm) ... breaks compatibility with older versions of NPLM, and older models.	2014-11-17 15:12:18 +00:00
Rico Sennrich	00c35fb565	fix LMOOVFeature and implement it for NeuralLMWrapper	2014-11-13 10:36:26 +00:00
Rico Sennrich	b42591d585	NPLM could implement shared or thread-local cache; wrapper works with either.	2014-10-16 14:33:35 +01:00
Rico Sennrich	b0fbf80dc0	remove unnecessary code, and floor NPLM scores	2014-07-18 14:39:54 +01:00
Rico Sennrich	eb5336ad9c	thread-safety, precomputation and caching for NeuralLMWrapper ... doesn't work with default nplm, but with the fork at https://github.com/rsennrich/nplm	2014-07-17 16:50:08 +01:00
Nicola Bertoldi	bd83999264	beautify	2014-01-15 16:42:02 +01:00
Hieu Hoang	a5f391e3f3	remove description argument from feature function constructor. Redundant	2013-10-29 18:59:53 +00:00
Hieu Hoang	518a766ed2	add NeuralLM	2013-10-28 23:34:50 +00:00
Hieu Hoang	37896ed854	add NeuralLM	2013-10-28 23:21:02 +00:00