1530ae4 | Kenneth Heafield | 09 January 2013, 13:15:04 UTC | Fix state comparison (impacted 32-bit) | 09 January 2013, 13:15:04 UTC |
936dbf6 | Barry Haddow | 08 January 2013, 16:40:00 UTC | Instance weighting | 08 January 2013, 16:40:00 UTC |
c86c11a | Barry Haddow | 08 January 2013, 15:34:29 UTC | instance weighting of lex weights | 08 January 2013, 15:34:29 UTC |
a55a936 | Barry Haddow | 08 January 2013, 14:28:16 UTC | remove warning | 08 January 2013, 14:28:16 UTC |
1b0f9f2 | Hieu Hoang | 08 January 2013, 13:54:35 UTC | re-enable some score regressions tests. Was failing due to rounding errors | 08 January 2013, 13:54:35 UTC |
5fe6868 | Hieu | 07 January 2013, 18:04:41 UTC | Merge branch 'master' of https://github.com/moses-smt/mosesdecoder | 07 January 2013, 18:04:41 UTC |
8e4f5a9 | Hieu | 07 January 2013, 18:03:40 UTC | eclipse | 07 January 2013, 18:03:40 UTC |
9973832 | Hieu | 07 January 2013, 18:03:32 UTC | eclipse | 07 January 2013, 18:03:32 UTC |
86327b1 | Hieu | 07 January 2013, 18:03:20 UTC | eclipse | 07 January 2013, 18:03:20 UTC |
3bf83af | Hieu | 07 January 2013, 18:03:11 UTC | eclipse | 07 January 2013, 18:03:11 UTC |
18b1622 | Hieu | 07 January 2013, 18:02:55 UTC | eclipse | 07 January 2013, 18:02:55 UTC |
133e375 | Hieu | 07 January 2013, 18:02:40 UTC | eclipse | 07 January 2013, 18:02:40 UTC |
366ca8a | Kenneth Heafield | 07 January 2013, 17:38:04 UTC | Remove if that doesn't do anything, define _FILE_OFFSET_BITS 64 | 07 January 2013, 17:38:04 UTC |
506cb8e | Hieu Hoang | 07 January 2013, 14:14:39 UTC | change lseek64() to lseek() and off64_t to off_t. Cygwin doesn't have those functions/types. They should be 64-bit if _FILE_OFFSET_BITS == 64 is set | 07 January 2013, 14:14:39 UTC |
ad209c8 | Kenneth Heafield | 04 January 2013, 22:42:32 UTC | Collapse lseek macro | 04 January 2013, 22:42:32 UTC |
77a3f32 | Hieu Hoang | 04 January 2013, 22:32:48 UTC | mac doesn't have lseek64() | 04 January 2013, 22:32:48 UTC |
f9ee7ae | Kenneth Heafield | 04 January 2013, 21:02:47 UTC | KenLM 0e5d259 including read_compressed fix | 04 January 2013, 21:02:47 UTC |
3203f7c | Kenneth Heafield | 04 January 2013, 21:02:13 UTC | If statically linking, check static libs for -mt | 04 January 2013, 21:02:13 UTC |
c8448de | Hieu Hoang | 04 January 2013, 20:00:23 UTC | Merge branch 'master' of git://github.com/moses-smt/mosesdecoder | 04 January 2013, 20:00:23 UTC |
43337df | Hieu Hoang | 04 January 2013, 20:00:04 UTC | get rid of HAVE_BOOST macro. Always has boost now | 04 January 2013, 20:00:04 UTC |
459acf8 | Barry Haddow | 04 January 2013, 14:55:24 UTC | Add support for instance weights file | 04 January 2013, 14:55:24 UTC |
1dfbe11 | hieu | 03 January 2013, 15:01:30 UTC | delete MergeScorer in mert/ | 03 January 2013, 15:01:30 UTC |
15c776e | hieu | 03 January 2013, 14:35:39 UTC | delete check and exit of Suffix Array phrase table implementation erroneously checked in | 03 January 2013, 14:35:39 UTC |
861792b | Barry Haddow | 21 December 2012, 15:39:25 UTC | extract can read an instance weights file. Still have to parallelise. | 21 December 2012, 15:39:25 UTC |
8fe900d | Barry Haddow | 21 December 2012, 12:35:11 UTC | Testing of phrase length feature | 21 December 2012, 12:35:11 UTC |
ce1b650 | Tetsuo Kiso | 20 December 2012, 18:06:41 UTC | Fix memory leak. The object was allocated with new, but it was not deleted. This may not be a serious problem because the program mostly runs a short time. However, it is not a good practice. | 20 December 2012, 18:06:41 UTC |
139148b | Phil Williams | 17 December 2012, 20:08:02 UTC | extract-ghkm and friends: don't unescape special characters Don't unescape special characters when reading XML parse trees in extract-ghkm, extract-rules, and relax-parse. | 17 December 2012, 20:08:02 UTC |
0ca5b89 | Phil Williams | 17 December 2012, 19:43:42 UTC | extract-ghkm: tweak label collection for unknown words Produce a better label set when unary rule elimination is enabled. | 17 December 2012, 19:43:42 UTC |
fb8d20a | Phil Williams | 17 December 2012, 19:02:30 UTC | extract-ghkm: --UnknownWordMinRelFreq, --UnknownWordUniform | 17 December 2012, 19:02:30 UTC |
06081f7 | Phil Williams | 17 December 2012, 18:49:50 UTC | extract-target-trees.py: minor fixes, code style | 17 December 2012, 18:49:50 UTC |
b275c94 | phikoehn | 12 December 2012, 07:02:59 UTC | allow for inclusion of extract from previous run | 12 December 2012, 07:02:59 UTC |
24e1df7 | phikoehn | 12 December 2012, 03:59:14 UTC | support for use of baseline alignment model | 12 December 2012, 03:59:14 UTC |
438dcb1 | phikoehn | 10 December 2012, 23:50:14 UTC | bug fix in experiment.perl wrt. get-corpus-script | 10 December 2012, 23:50:14 UTC |
16ea68f | Barry Haddow | 10 December 2012, 15:54:24 UTC | Fix bug in mml scoring Line length calculation was out of step with LM scoring. | 10 December 2012, 15:54:24 UTC |
598d65b | Ales Tamchyna | 10 December 2012, 12:12:34 UTC | adding a simple command-line utility for computing sentence-level BLEU (+1) | 10 December 2012, 12:12:34 UTC |
ed2d191 | phikoehn | 10 December 2012, 05:56:51 UTC | allow specification of end point for experiment.perl | 10 December 2012, 05:56:51 UTC |
ccf9e13 | phikoehn | 09 December 2012, 22:27:02 UTC | bug fix with multicore parallelizer | 09 December 2012, 22:27:02 UTC |
466b502 | phikoehn | 09 December 2012, 20:31:20 UTC | minor bug fixes with MML | 09 December 2012, 20:31:20 UTC |
2a3c9fc | Tetsuo Kiso | 06 December 2012, 23:45:47 UTC | Further optimization for extractor. Fixes inefficient updating N-gram counts. NOTE: Using '--binary' option (this option is not enabled by default yet) for saving outputs would lead to significant speed up. | 06 December 2012, 23:45:47 UTC |
8fdec9b | Tetsuo Kiso | 06 December 2012, 20:12:24 UTC | Use boost::unordered_map instead of std::map. For storing the word vocabulary used in computation of BLEU scores. This change will reduce the running time of extractor about 2-3 seconds (9% reduction). | 06 December 2012, 20:12:24 UTC |
6c04c4a | Tetsuo Kiso | 06 December 2012, 17:46:59 UTC | Add more tests to the Data class. | 06 December 2012, 17:46:59 UTC |
c7f6e38 | Tetsuo Kiso | 06 December 2012, 17:39:02 UTC | Use FilePiece to load N-best lists. Since FilePiece is friendly with StringPiece. | 06 December 2012, 17:39:02 UTC |
38e145e | Tetsuo Kiso | 06 December 2012, 16:39:22 UTC | Use util::TokenIter to tokenize n-best lists. Reduce creating std::string objects, too. In both ScoreArray and FeatureArray classes, the private members to track sentence indices (namely, "m_index") were unnecessarily declared as std::string, but it's better to directly declare them as 'int'. | 06 December 2012, 16:39:22 UTC |
cd3fb3b | Tetsuo Kiso | 06 December 2012, 14:46:22 UTC | Untabify. | 06 December 2012, 14:46:22 UTC |
ac045a1 | Tetsuo Kiso | 06 December 2012, 13:08:33 UTC | Speed up N-gram counts when running extractor. By replacing std::map with boost::unordered_map. Runtime of extractor on 100-best lists of 2679 sentences: Before: real 0m35.314s user 0m34.030s sys 0m1.280s Ater: real 0m26.729s user 0m25.420s sys 0m1.310s | 06 December 2012, 13:08:33 UTC |
55f65c3 | Hieu Hoang | 03 December 2012, 14:57:33 UTC | race condition in chart decoding with -T arg | 03 December 2012, 14:57:33 UTC |
ab2effb | phikoehn | 01 December 2012, 13:46:59 UTC | train MML in-/out-of-domain language models with same vocabulary | 01 December 2012, 13:46:59 UTC |
269883f | phikoehn | 01 December 2012, 13:45:00 UTC | Merge branch 'master' of git://github.com/moses-smt/mosesdecoder | 01 December 2012, 13:45:00 UTC |
0c5d000 | phikoehn | 01 December 2012, 13:44:57 UTC | my change to weight-wt | 01 December 2012, 13:44:57 UTC |
205cea8 | Marcin Junczys-Dowmunt | 30 November 2012, 23:38:20 UTC | Allow .minlexr suffix and bugfix | 30 November 2012, 23:38:20 UTC |
650d31f | Eva Hasler | 30 November 2012, 18:04:50 UTC | don't need to specify weight-wt | 30 November 2012, 18:04:50 UTC |
a07f71d | Hieu Hoang | 30 November 2012, 17:15:32 UTC | race condition on letter sed cache. Requires locking | 30 November 2012, 17:15:32 UTC |
7abb3c8 | Hieu Hoang | 30 November 2012, 13:50:59 UTC | remove locking. Make wordIndex variable local | 30 November 2012, 13:50:59 UTC |
5fd9cbb | Hieu Hoang | 30 November 2012, 10:28:51 UTC | delete reference to numpy. Doesn't need it | 30 November 2012, 10:28:51 UTC |
017bbe7 | Hieu Hoang | 30 November 2012, 09:49:36 UTC | forgotten misc programs for Compact pt | 30 November 2012, 09:49:36 UTC |
338b765 | phikoehn | 30 November 2012, 07:36:59 UTC | ooops | 30 November 2012, 07:36:59 UTC |
84cb04c | phikoehn | 30 November 2012, 07:28:31 UTC | fixes and extensions to modified Moore-Lewis filtering, now works with domain features | 30 November 2012, 07:28:31 UTC |
1f7ee0e | phikoehn | 29 November 2012, 23:44:10 UTC | change of settings for sigtest filtering | 29 November 2012, 23:44:10 UTC |
d4ead15 | Hieu Hoang | 29 November 2012, 15:27:38 UTC | fuzzy match phrase-table is multi-threaded | 29 November 2012, 15:27:38 UTC |
9aad7c6 | Hieu Hoang | 27 November 2012, 18:04:01 UTC | move CompactPt to TranslationModel/ | 27 November 2012, 18:04:01 UTC |
1520640 | Hieu Hoang | 27 November 2012, 17:33:42 UTC | Merge branch 'master' of github.com:moses-smt/mosesdecoder | 27 November 2012, 17:33:42 UTC |
b317ac1 | Hieu Hoang | 27 November 2012, 17:33:04 UTC | compile error on misc programs | 27 November 2012, 17:33:04 UTC |
bc1e967 | Hieu Hoang | 27 November 2012, 17:23:31 UTC | move CKY+Parser to TranslationModel/ | 27 November 2012, 17:23:31 UTC |
ae8a48b | Hieu Hoang | 27 November 2012, 17:09:23 UTC | move Score3Parser to TranslationModel/ | 27 November 2012, 17:09:23 UTC |
1aae9aa | Hieu Hoang | 27 November 2012, 16:57:23 UTC | move RuleTable to TranslationModel/ | 27 November 2012, 16:57:23 UTC |
6bf2870 | Hieu Hoang | 27 November 2012, 16:31:42 UTC | move the rest of DynSA to TranslationModel/ | 27 November 2012, 16:31:42 UTC |
4d8e4ae | Hieu Hoang | 27 November 2012, 16:16:30 UTC | move DynSAInclude to TranslationModel/ | 27 November 2012, 16:16:30 UTC |
f0e1291 | Barry Haddow | 27 November 2012, 15:58:55 UTC | mml-score.py. Support for combining with domain features. | 27 November 2012, 15:58:55 UTC |
75108c0 | Hieu Hoang | 27 November 2012, 15:39:08 UTC | minor debug messages | 27 November 2012, 15:39:08 UTC |
0b54d32 | Hieu Hoang | 27 November 2012, 15:36:24 UTC | move fuzzy-match to TranslationModel/ | 27 November 2012, 15:36:24 UTC |
59449f2 | Hieu Hoang | 27 November 2012, 15:08:31 UTC | make TranslationModel subdirectory and move files from moses/ into it | 27 November 2012, 15:08:31 UTC |
a4cbaa5 | Hieu Hoang | 27 November 2012, 13:17:44 UTC | multithreaded fuzzy match | 27 November 2012, 13:17:44 UTC |
b30b2ff | Hieu Hoang | 27 November 2012, 12:29:36 UTC | compile error in fuzzy match | 27 November 2012, 12:29:36 UTC |
9e08ef8 | Hieu Hoang | 27 November 2012, 10:42:28 UTC | multithreaded fuzzy match | 27 November 2012, 10:42:28 UTC |
ff1b210 | Hieu Hoang | 26 November 2012, 23:10:40 UTC | Merge branch 'master' of github.com:moses-smt/mosesdecoder | 26 November 2012, 23:10:40 UTC |
ea2ad97 | Hieu Hoang | 26 November 2012, 23:10:23 UTC | eclipse files | 26 November 2012, 23:10:23 UTC |
51afd32 | phikoehn | 24 November 2012, 20:13:30 UTC | Merge branch 'master' of git://github.com/moses-smt/mosesdecoder | 24 November 2012, 20:13:30 UTC |
b5d0874 | phikoehn | 24 November 2012, 20:13:14 UTC | extensions to modified moore-lewis filtering, bug fixes | 24 November 2012, 20:13:14 UTC |
bb76c15 | marcinj | 24 November 2012, 16:58:07 UTC | Added another consistency check for false positives | 24 November 2012, 16:58:07 UTC |
ea610a0 | phikoehn | 24 November 2012, 12:43:13 UTC | added modified-moore-lewis from Barry Haddow into EMS | 24 November 2012, 12:43:13 UTC |
7b36e11 | Hieu Hoang | 23 November 2012, 18:52:06 UTC | make fuzzy match pt threadable by marking it as non-threadable! TODO - do it properly | 23 November 2012, 18:52:06 UTC |
cb95548 | Hieu Hoang | 23 November 2012, 17:38:57 UTC | Merge branch 'master' of github.com:moses-smt/mosesdecoder | 23 November 2012, 17:38:57 UTC |
d4cebb0 | phikoehn | 23 November 2012, 17:35:13 UTC | added ems support for sigtest-filter | 23 November 2012, 17:35:13 UTC |
65457aa | Hieu Hoang | 23 November 2012, 17:33:00 UTC | load source corpus along with suffix array. Use vocab from SA execlusively | 23 November 2012, 17:33:00 UTC |
d0cf8f4 | Hieu Hoang | 22 November 2012, 17:37:36 UTC | order of lexical probability has flipped | 22 November 2012, 17:37:36 UTC |
487822e | Hieu Hoang | 22 November 2012, 15:08:00 UTC | don't write to stdout | 22 November 2012, 15:08:00 UTC |
b299702 | Hieu Hoang | 21 November 2012, 16:41:05 UTC | add test misc perl file | 21 November 2012, 16:41:05 UTC |
c2a96fc | phikoehn | 20 November 2012, 17:19:17 UTC | adjust to irstlm changes | 20 November 2012, 17:19:17 UTC |
5cd614e | phikoehn | 20 November 2012, 17:18:57 UTC | adjust to irstlm changes | 20 November 2012, 17:18:57 UTC |
7d6d91a | Hieu Hoang | 20 November 2012, 16:42:49 UTC | move zmert to contrib folder | 20 November 2012, 16:42:49 UTC |
feff7af | Hieu Hoang | 20 November 2012, 16:14:20 UTC | add tests for decoding with compressed pt | 20 November 2012, 16:14:20 UTC |
dbf0123 | Hieu Hoang | 20 November 2012, 13:22:03 UTC | don't load compact lex model if it doesn't exist | 20 November 2012, 13:22:03 UTC |
0bf0666 | Hieu Hoang | 20 November 2012, 12:11:01 UTC | add error message for non-existant input file | 20 November 2012, 12:11:01 UTC |
83d0797 | Hieu Hoang | 19 November 2012, 19:44:12 UTC | add tests for decoding with compressed pt. Fails | 19 November 2012, 19:44:12 UTC |
b7272b4 | Kenneth Heafield | 19 November 2012, 17:59:33 UTC | BINDIR and LIBDIR constants | 19 November 2012, 17:59:33 UTC |
da39cff | Kenneth Heafield | 18 November 2012, 14:29:43 UTC | D'oh <= | 18 November 2012, 14:29:43 UTC |
f09fb75 | Marcin Junczys-Dowmunt | 18 November 2012, 09:21:35 UTC | Compact reordering table, do not force full recompilation | 18 November 2012, 09:21:35 UTC |
e543608 | Marcin Junczys-Dowmunt | 17 November 2012, 21:03:49 UTC | Do not search for compact reordering table if CMPH is not compile into moses | 17 November 2012, 21:03:49 UTC |
28a962e | marcinj | 17 November 2012, 17:25:30 UTC | Change -alignment-info to -no-alignment-info | 17 November 2012, 17:25:30 UTC |
79526d0 | marcinj | 17 November 2012, 09:07:33 UTC | more consistent file name resolution | 17 November 2012, 09:07:33 UTC |