Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/moses-smt/mosesdecoder.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/vw
diff options
context:
space:
mode:
authorMarcin Junczys-Dowmunt <junczys@amu.edu.pl>2015-01-09 17:13:59 +0300
committerMarcin Junczys-Dowmunt <junczys@amu.edu.pl>2015-01-09 17:13:59 +0300
commitb05f97b5d124fe756472a3adf0e696c093467b53 (patch)
treee7e79f6d878aeb687ec04ec2e3c83c21acba92bf /vw
parent685b6574cceaa12b06606b0a0ac835353b7bc064 (diff)
Update README.md
Diffstat (limited to 'vw')
-rw-r--r--vw/README.md15
1 files changed, 8 insertions, 7 deletions
diff --git a/vw/README.md b/vw/README.md
index 0d053d7b5..29f1bb861 100644
--- a/vw/README.md
+++ b/vw/README.md
@@ -7,13 +7,14 @@ function.
Implemented classifier features
-------------------------------
-* VWFeatureSourceBagOfWords: This create a feature of form bow^token for every
+* VWFeatureSourceBagOfWords: This creates a feature of form bow^token for every
source sentence token.
-* VWFeatureSourceExternalFeatures:
-* VWFeatureSourcePhraseInternal:
-* VWFeatureSourceWindow:
-* VWFeatureTargetIndicator:
-* VWFeatureTargetPhraseInternal:
+* VWFeatureSourceExternalFeatures: (not quite finished yet) when used with -inputtype 5 this can be used to supply additional feature to VW. The input is a tab-separated file, the first column is usual input sentence, all other columns can be used for meta-data. By default the second column is read in and features are split on whitespace.
+* VWFeatureSourceIndicator: Ass a feature for the whole source phrase.
+* VWFeatureSourcePhraseInternal: Adds a separate feature for every word of the source phrase.
+* VWFeatureSourceWindow: Adds source words in a window before and after the source phrase as features. These does not overlap with VWFeatureSourcePhraseInternal.
+* VWFeatureTargetIndicator: Adds a feature for the whole target phrase.
+* VWFeatureTargetPhraseInternal: Adds a separate feature for every word of the target phrase.
Configuration
-------------
@@ -69,4 +70,4 @@ You can also use multiple classifiers:
...
Training the classifier
------------------------ \ No newline at end of file
+-----------------------