Age | Commit message (Collapse) | Author | |
---|---|---|---|
2013-11-05 | cleaning up syntax features | Patrick Simianer | |
2013-04-29 | enabled build | Patrick Simianer | |
2013-04-29 | added (broken) variant of a feature | Patrick Simianer | |
2013-04-26 | added Gesa's alternative source syntax features | Patrick Simianer | |
2013-04-26 | added (broken) feature function | Patrick Simianer | |
2013-04-26 | added Gesa's soft syntactic features | Patrick Simianer | |
2013-04-25 | added Gesa's parse match feature function | Patrick Simianer | |
2013-03-14 | fix source path | Chris Dyer | |
2013-03-14 | source path features | Chris Dyer | |
2012-11-05 | merge upstream/master | Patrick Simianer | |
2012-10-16 | clean up of bad header includes | Chris Dyer | |
2012-10-11 | add support for dlopen'd feature functions | Chris Dyer | |
2012-06-24 | RuleTargetBigramFeatures, parallelize.rb | Patrick Simianer | |
2012-03-08 | simple context feature for tagger | Chris Dyer | |
2011-10-14 | remove FSA integration code. will have to be resurrected another day | Chris Dyer | |
2011-09-17 | enable sss features | Chris Dyer | |
2011-09-12 | source syntax features ~ blunsom emnlp 2008 | Chris Dyer | |
2011-09-09 | rule feature refactoring | Chris Dyer | |
2011-08-18 | KenLM update: Bhiksha's trick, simple test for lms without unk, auto-detect ↵ | Kenneth Heafield | |
binary files instead of requiring them to be specified at runtime. | |||
2011-07-06 | ngram count features | Chris Dyer | |
2011-06-26 | Quantization | Kenneth Heafield | |
2011-06-01 | register feature | Chris Dyer | |
2011-05-20 | kenlm update including being nicer to NFS | Kenneth Heafield | |
2011-04-22 | fix | Chris Dyer | |
2011-04-22 | make compatible with FastSparseVector | Chris Dyer | |
2011-02-26 | chiang, marton, resnik fine reordering features | Chris Dyer | |
2011-02-02 | first span features | Chris Dyer | |
2010-12-28 | incorporate dwarf features | chris dyer | |
2010-12-23 | non-latin character detector | Chris Dyer | |
2010-12-23 | support different types in kenlm | Chris Dyer | |
2010-12-22 | clean up names of feature functions, fix tagger, fix tests | Chris Dyer | |
2010-12-15 | feature to detect self-transition before/after identity translations | Chris Dyer | |
2010-12-09 | major refactor of markov features for word alignment | Chris Dyer | |
2010-12-01 | alternative def of neighborhoods | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@739 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-11-12 | Gross hack for compiling GLC as a library. Currently has hardcoded path, but ↵ | jon.h.clark | |
shouldn't break the build. git-svn-id: https://ws10smt.googlecode.com/svn/trunk@716 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-11-11 | klm stub | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@711 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-10-21 | bit more alignment stuff | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@686 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-10-13 | target unigram feature | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@672 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-10-13 | trigger ff, max iteration for online optimizer | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@671 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-09-29 | another feature, another POS | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@664 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-09-19 | big refactor | redpony | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@649 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-08-07 | cdec -A "LanguageModelFsa lm.gz" works | graehl@gmail.com | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@490 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-08-07 | propagation of feature name+debug from factory, return correct features ↵ | graehl | |
array for fsa ffs git-svn-id: https://ws10smt.googlecode.com/svn/trunk@486 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-08-07 | usage for FSA features | graehl | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@485 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-08-07 | dynamic fsa ff, factory for fsa and ff shares code, factory moved to ↵ | graehl | |
ff_factory.cc git-svn-id: https://ws10smt.googlecode.com/svn/trunk@483 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-07-27 | sample fsa ffs compile | graehl | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@432 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-07-27 | fsa feature templated Accum interface, phrase interface allows exceeding ↵ | graehl | |
markov order e.g. unigram state, 3gram lm. use Accum,set_value rather than clear,add_value. warning: 3gram fsa lm disagrees with bottom-up in 4th decimal place git-svn-id: https://ws10smt.googlecode.com/svn/trunk@431 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-07-26 | LanguageModelFsa works. TODO: sri context shortening? | graehl | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@414 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-07-23 | fsa: stateless works, debug sample bigram {Longer,Shorter}ThanPrev | graehl | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@375 ec762483-ff6d-05da-a07a-a48fb63a330f | |||
2010-07-23 | bottom-up FF from fsa FF - WordPenaltyFsa - needs debugging | graehl | |
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@373 ec762483-ff6d-05da-a07a-a48fb63a330f |