moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2025-01-02 17:09:36 +03:00
Code Issues Packages Projects Releases Wiki Activity

EMS: add TRAINING:no-glue-grammar option

Browse Source
This commit is contained in:
Phil Williams 2015-03-03 12:36:09 +00:00
parent 2543c78b71
commit 90e8d4940c
2 changed files with 20 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
scripts/ems/experiment.meta
View File

@ -644,7 +644,7 @@ create-config
in: sigtest-filter-reordering-table sigtest-filter-phrase-translation-table transliteration-table generation-table sparse corpus-mml-prefilter=OR=corpus-mml-postfilter=OR=domains osm-model INTERPOLATED-LM:binlm LM:binlm
out: config
ignore-if: use-hiero thot
rerun-on-change: decoding-steps alignment-factors translation-factors reordering-factors generation-factors lexicalized-reordering training-options script decoding-graph-backoff score-settings additional-ini mmsapt
rerun-on-change: decoding-steps alignment-factors translation-factors reordering-factors generation-factors lexicalized-reordering training-options script decoding-graph-backoff score-settings additional-ini mmsapt no-glue-grammar
default-name: model/moses.ini
error: Unknown option
final-model: yes
@ -946,14 +946,14 @@ filter
in: input TRAINING:sigtest-filter-phrase-translation-table TRAINING:sigtest-filter-reordering-table TRAINING:corpus-mml-prefilter=OR=TRAINING:corpus-mml-postfilter=OR=TRAINING:domains TRAINING:transliteration-table
out: filtered-dir
default-name: tuning/filtered
rerun-on-change: filter-settings ttable-binarizer
rerun-on-change: filter-settings ttable-binarizer TRAINING:no-glue-grammar
ignore-if: TRAINING:binarize-all
error: already exists. Please delete
filter-devtest
in: input-devtest TRAINING:sigtest-filter-phrase-translation-table TRAINING:sigtest-filter-reordering-table
out: filtered-dir-devtest
default-name: tuning/filtered.devtest
rerun-on-change: filter-settings ttable-binarizer
rerun-on-change: filter-settings ttable-binarizer TRAINING:no-glue-grammar
ignore-if: TRAINING:binarize-all
ignore-unless: use-mira
error: already exists. Please delete
@ -1089,7 +1089,7 @@ filter
in: input TRAINING:sigtest-filter-phrase-translation-table TRAINING:sigtest-filter-reordering-table TRAINING:corpus-mml-prefilter=OR=TRAINING:corpus-mml-postfilter=OR=TRAINING:domains TRAINING:transliteration-table
out: filtered-dir
default-name: evaluation/filtered
rerun-on-change: filter-settings report-precision-by-coverage ttable-binarizer
rerun-on-change: filter-settings report-precision-by-coverage ttable-binarizer TRAINING:no-glue-grammar
pass-if: TRAINING:binarize-all
ignore-if: use-hiero
error: already exists. Please delete

25
scripts/ems/experiment.perl
View File

@ -2206,10 +2206,13 @@ sub define_training_extract_phrases {
$cmd .= "-corpus $corpus ";
if (&get("TRAINING:hierarchical-rule-set")) {
my $glue_grammar_file = &get("TRAINING:glue-grammar");
$glue_grammar_file = &versionize(&long_file_name("glue-grammar","model",""))
unless $glue_grammar_file;
$cmd .= "-glue-grammar-file $glue_grammar_file ";
my $no_glue_grammar = &get("TRAINING:no-glue-grammar");
if (!defined($no_glue_grammar) || $no_glue_grammar eq "false") {
my $glue_grammar_file = &get("TRAINING:glue-grammar");
$glue_grammar_file = &versionize(&long_file_name("glue-grammar","model",""))
unless $glue_grammar_file;
$cmd .= "-glue-grammar-file $glue_grammar_file ";
}
if (&get("GENERAL:output-parser") && (&get("TRAINING:use-unknown-word-labels") || &get("TRAINING:use-unknown-word-soft-matches"))) {
my $unknown_word_label = &versionize(&long_file_name("unknown-word-label","model",""));
@ -2418,10 +2421,13 @@ sub get_config_tables {
if (&get("TRAINING:hierarchical-rule-set")) {
$extract_version = $RE_USE[$STEP_LOOKUP{"TRAINING:extract-phrases"}]
if defined($STEP_LOOKUP{"TRAINING:extract-phrases"});
my $glue_grammar_file = &get("TRAINING:glue-grammar");
$glue_grammar_file = &versionize(&long_file_name("glue-grammar","model",""),$extract_version)
unless $glue_grammar_file;
$cmd .= "-glue-grammar-file $glue_grammar_file ";
my $no_glue_grammar = &get("TRAINING:no-glue-grammar");
if (!defined($no_glue_grammar) || $no_glue_grammar eq "false") {
my $glue_grammar_file = &get("TRAINING:glue-grammar");
$glue_grammar_file = &versionize(&long_file_name("glue-grammar","model",""),$extract_version)
unless $glue_grammar_file;
$cmd .= "-glue-grammar-file $glue_grammar_file ";
}
}
# additional settings for syntax models
@ -2721,6 +2727,7 @@ sub get_training_setting {
my $parallel = &get("TRAINING:parallel");
my $pcfg = &get("TRAINING:use-pcfg-feature");
my $baseline_alignment = &get("TRAINING:baseline-alignment-model");
my $no_glue_grammar = &get("TRAINING:no-glue-grammar");
my $xml = $source_syntax || $target_syntax;
@ -2740,7 +2747,7 @@ sub get_training_setting {
$cmd .= "-xml " if $xml;
$cmd .= "-target-syntax " if $target_syntax;
$cmd .= "-source-syntax " if $source_syntax;
$cmd .= "-glue-grammar " if $hierarchical;
$cmd .= "-glue-grammar " if $hierarchical && (!defined($no_glue_grammar) || $no_glue_grammar eq "false");
$cmd .= "-score-options '".$score_settings."' " if $score_settings;
$cmd .= "-parallel " if $parallel;
$cmd .= "-pcfg " if $pcfg;