You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@joshua.apache.org by mj...@apache.org on 2016/06/23 18:46:01 UTC
[50/60] [partial] incubator-joshua git commit: maven multi-module
layout 1st commit: moving files into joshua-core
http://git-wip-us.apache.org/repos/asf/incubator-joshua/blob/e2734396/joshua-core/resources/kbest_extraction/grammar
----------------------------------------------------------------------
diff --git a/joshua-core/resources/kbest_extraction/grammar b/joshua-core/resources/kbest_extraction/grammar
new file mode 100644
index 0000000..a03b2d9
--- /dev/null
+++ b/joshua-core/resources/kbest_extraction/grammar
@@ -0,0 +1,25 @@
+[X] ||| a ||| A ||| 2
+[X] ||| a ||| B ||| 3
+[X] ||| a ||| C ||| 5
+[X] ||| a ||| D ||| 7
+[X] ||| a ||| E ||| 11
+[X] ||| b ||| A ||| 13
+[X] ||| b ||| B ||| 17
+[X] ||| b ||| C ||| 19
+[X] ||| b ||| D ||| 23
+[X] ||| b ||| E ||| 29
+[X] ||| c ||| A ||| 31
+[X] ||| c ||| B ||| 37
+[X] ||| c ||| C ||| 41
+[X] ||| c ||| D ||| 43
+[X] ||| c ||| E ||| 47
+[X] ||| d ||| A ||| 53
+[X] ||| d ||| B ||| 59
+[X] ||| d ||| C ||| 61
+[X] ||| d ||| D ||| 67
+[X] ||| d ||| E ||| 71
+[X] ||| e ||| A ||| 73
+[X] ||| e ||| B ||| 79
+[X] ||| e ||| C ||| 83
+[X] ||| e ||| D ||| 89
+[X] ||| e ||| E ||| 97
http://git-wip-us.apache.org/repos/asf/incubator-joshua/blob/e2734396/joshua-core/resources/kbest_extraction/joshua.config
----------------------------------------------------------------------
diff --git a/joshua-core/resources/kbest_extraction/joshua.config b/joshua-core/resources/kbest_extraction/joshua.config
new file mode 100644
index 0000000..cdab98e
--- /dev/null
+++ b/joshua-core/resources/kbest_extraction/joshua.config
@@ -0,0 +1,27 @@
+feature-function = StateMinimizingLanguageModel -lm_type kenlm -lm_order 5 -lm_file resources/kbest_extraction/lm.gz
+
+tm = thrax -owner pt -maxspan 12 -path resources/kbest_extraction/grammar
+tm = thrax -owner glue -maxspan -1 -path resources/kbest_extraction/glue-grammar
+
+mark_oovs=false
+
+#tm config
+default_non_terminal=X
+goalSymbol=GOAL
+
+#pruning config
+pop-limit=100
+
+#nbest config
+use_unique_nbest=true
+top-n = 3126
+
+#feature_function = WordPenalty
+feature_function = OOVPenalty
+
+# Model Weights ####
+
+lm_0 1
+tm_pt_0 1
+tm_glue_0 1
+OOVPenalty 10000
http://git-wip-us.apache.org/repos/asf/incubator-joshua/blob/e2734396/joshua-core/resources/kbest_extraction/lm.gz
----------------------------------------------------------------------
diff --git a/joshua-core/resources/kbest_extraction/lm.gz b/joshua-core/resources/kbest_extraction/lm.gz
new file mode 100644
index 0000000..a26335e
Binary files /dev/null and b/joshua-core/resources/kbest_extraction/lm.gz differ