Rico Sennrich
|
6748f4106f
|
compatibility with NPLM 0.3 (breaks earlier versions of NPLM)
|
2014-11-17 18:08:42 +00:00 |
|
XapaJIaMnu
|
a343837095
|
Add option to choose activation function during nplm training
|
2014-11-15 11:54:47 +00:00 |
|
Paul Baltescu
|
35401f7dd3
|
Merge branch 'bilingual-lm' of github.com:moses-smt/mosesdecoder into bilingual-lm
|
2014-11-14 12:39:02 +00:00 |
|
Paul Baltescu
|
fefd2b0ada
|
Fix merge errors.
|
2014-11-13 16:52:24 +00:00 |
|
XapaJIaMnu
|
2028b86642
|
BilingualNPLM requries specific source and target vocabulary lists
|
2014-11-13 16:14:17 +00:00 |
|
XapaJIaMnu
|
d5567b6cfb
|
Training: Do the preparation step ourselves. No validation support yet. No decoder support yet.
|
2014-11-13 16:14:17 +00:00 |
|
Rico Sennrich
|
9b161dc888
|
floor score of bilingualNPLM
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
b254fcea99
|
moses/LM/BilingualLM.cpp
|
2014-11-13 16:14:16 +00:00 |
|
Rico Sennrich
|
8fd3be9e4e
|
add EOS token </s> to each sentence
|
2014-11-13 16:14:16 +00:00 |
|
Rico Sennrich
|
f26fc251d5
|
sort vocab by frequency
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
bb70f60f67
|
grrr
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
e330ab35d5
|
Short option must be only one letter
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
a74105ea7d
|
Fix a wrong condition
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
e54c171850
|
Make it optional to prepare the validation set
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
a300824bd1
|
Add optional validation during training
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
f6c64adb92
|
Remove unnecessary const gimmicks
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
18a6d12cb0
|
Rework lookup and greatly speedup decoding (2x+)
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
cf3fe60cf6
|
Backoff to null token in hiero
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
06fa9b5916
|
Add option to build with Og for newer versions of gcc
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
493f05eb97
|
Remove old unused code from BilingualLM
|
2014-11-13 16:14:16 +00:00 |
|
Paul Baltescu
|
211878c9d3
|
Make OxLM extensions compile.
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
8e4ff790a7
|
Add null word support for hiero.
|
2014-11-13 16:14:16 +00:00 |
|
Paul Baltescu
|
0578a31799
|
Flag for setting OxLM models to unnormalized.
|
2014-11-13 16:14:16 +00:00 |
|
XapaJIaMnu
|
0451142ece
|
Add null token normalization for models to be used with the chart decoder.
|
2014-11-13 16:13:38 +00:00 |
|
XapaJIaMnu
|
aae894fe6b
|
Add null token in vocabulary during construction
|
2014-11-13 16:13:38 +00:00 |
|
XapaJIaMnu
|
b4f51c05d1
|
Add option to reduce the ngrams from already prepared .ngrams file to train a model with smaller number of ngrams
|
2014-11-13 16:13:38 +00:00 |
|
XapaJIaMnu
|
5851a2c2bb
|
Prevent realocation of vectors
|
2014-11-13 16:13:38 +00:00 |
|
Paul Baltescu
|
2705a47876
|
Fix OxLM.
|
2014-11-13 16:13:38 +00:00 |
|
XapaJIaMnu
|
02c375ef78
|
Refactor the BilingualLM for chart to make it faster. Untested
|
2014-11-13 16:13:04 +00:00 |
|
XapaJIaMnu
|
7858f74e9e
|
Rename BilingualLM_NPLM so that it is not confused with a sparse feature
|
2014-11-13 16:13:04 +00:00 |
|
Paul Baltescu
|
61826cee8a
|
Rename OxLM features.
|
2014-11-13 16:13:04 +00:00 |
|
Paul Baltescu
|
32c169c25f
|
Optional back-off to POS tags in OxLM.
|
2014-11-13 16:12:19 +00:00 |
|
Paul Baltescu
|
86d64b65e2
|
Correctly map the source unknown token.
|
2014-11-13 16:10:40 +00:00 |
|
Paul Baltescu
|
167e272818
|
Fix Bilingual OxLM context word order.
|
2014-11-13 16:10:40 +00:00 |
|
Paul Baltescu
|
5f9d481ee6
|
Make query cache sentence specific.
|
2014-11-13 16:10:40 +00:00 |
|
Paul Baltescu
|
248aa4bf8a
|
Convert moses words to oxlm word ids.
|
2014-11-13 16:10:40 +00:00 |
|
Paul Baltescu
|
97b632e045
|
Clean up OxLMMapper.
|
2014-11-13 16:08:56 +00:00 |
|
Paul Baltescu
|
90ebf13789
|
Set BilingualLM parameters nicely.
|
2014-11-13 16:08:10 +00:00 |
|
Paul Baltescu
|
7588c4b8e3
|
Skeleton for source conditioned OxLM feature.
|
2014-11-13 16:08:10 +00:00 |
|
Paul Baltescu
|
6f9d59129f
|
Rename LBLLM -> OxLM.
|
2014-11-13 16:07:38 +00:00 |
|
Paul Baltescu
|
af28063e3b
|
Fix compilation errors introduced by new oxlm changes.
|
2014-11-13 15:54:54 +00:00 |
|
Paul Baltescu
|
4811701277
|
Fix broken include.
|
2014-11-13 15:51:48 +00:00 |
|
Paul Baltescu
|
cb7167f088
|
Fix bugs in BilingualLM for chart based decoding.
|
2014-11-13 15:51:48 +00:00 |
|
Paul Baltescu
|
3624bd776c
|
Fix a few bugs in BilingualLM for phrase based decoding.
|
2014-11-13 15:51:48 +00:00 |
|
Paul Baltescu
|
5f87cf94d8
|
Move BilingualLM under LM.
|
2014-11-13 15:51:48 +00:00 |
|
XapaJIaMnu
|
fbac0ae418
|
Make sure we always have unk in the vocabulary, otherwise we get off-by-one indexes during decoding
|
2014-11-13 15:51:48 +00:00 |
|
XapaJIaMnu
|
961578286f
|
Forgot to close a file...
|
2014-11-13 15:51:48 +00:00 |
|
XapaJIaMnu
|
1bac666e5f
|
Fix small oversights
|
2014-11-13 15:51:48 +00:00 |
|
XapaJIaMnu
|
617ef015df
|
Extend train_nplm with various options
|
2014-11-13 15:51:48 +00:00 |
|
XapaJIaMnu
|
a1a10a9209
|
Remove unused variable that likely causes crashes
|
2014-11-13 15:51:48 +00:00 |
|