mirror of
https://github.com/moses-smt/mosesdecoder.git
synced 2024-12-25 12:52:29 +03:00
Update transliteration scripts to use the on-disk phrase table
The scripts now use CreateOnDiskPt instead of processPhraseTable (which
is no longer supported and was removed by commit f3a84fc01
).
This commit is contained in:
parent
fa3f82a04d
commit
7e54e23fe2
@ -120,7 +120,7 @@ sub run_transliteration
|
||||
|
||||
`$MOSES_SRC/scripts/training/train-model.perl -mgiza -mgiza-cpus 10 -dont-zip -first-step 9 -external-bin-dir $EXTERNAL_BIN_DIR -f $INPUT_EXTENSION -e $OUTPUT_EXTENSION -alignment grow-diag-final-and -parts 5 -score-options '--KneserNey' -phrase-translation-table $TRANSLIT_MODEL/model/phrase-table -config $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini -lm 0:3:$TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini:8`;
|
||||
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $TRANSLIT_MODEL/evaluation/$eval_file.filtered $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini $TRANSLIT_MODEL/evaluation/$eval_file -Binarizer "$MOSES_SRC/bin/processPhraseTable"`;
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $TRANSLIT_MODEL/evaluation/$eval_file.filtered $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini $TRANSLIT_MODEL/evaluation/$eval_file -Binarizer "$MOSES_SRC/bin/CreateOnDiskPt 1 1 4 100 2"`;
|
||||
|
||||
`rm $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini`;
|
||||
|
||||
|
@ -135,7 +135,7 @@ sub run_transliteration
|
||||
|
||||
`$MOSES_SRC/scripts/training/train-model.perl -mgiza -mgiza-cpus 10 -dont-zip -first-step 9 -external-bin-dir $EXTERNAL_BIN_DIR -f $INPUT_EXTENSION -e $OUTPUT_EXTENSION -alignment grow-diag-final-and -parts 5 -score-options '--KneserNey' -phrase-translation-table $TRANSLIT_MODEL/model/phrase-table -config $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini -lm 0:3:$TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini:8`;
|
||||
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $TRANSLIT_MODEL/evaluation/$eval_file.filtered $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini $TRANSLIT_MODEL/evaluation/$eval_file -Binarizer "$MOSES_SRC/bin/processPhraseTable"`;
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $TRANSLIT_MODEL/evaluation/$eval_file.filtered $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini $TRANSLIT_MODEL/evaluation/$eval_file -Binarizer "$MOSES_SRC/bin/CreateOnDiskPt 1 1 4 100 2"`;
|
||||
|
||||
`rm $TRANSLIT_MODEL/evaluation/$eval_file.moses.table.ini`;
|
||||
|
||||
@ -296,7 +296,7 @@ sub run_decoder
|
||||
|
||||
`$MOSES_SRC/scripts/training/train-model.perl -mgiza -mgiza-cpus 10 -dont-zip -first-step 9 -external-bin-dir $EXTERNAL_BIN_DIR -f $INPUT_EXTENSION -e $OUTPUT_EXTENSION -alignment grow-diag-final-and -parts 5 -lmodel-oov-feature "yes" -post-decoding-translit "yes" -phrase-translation-table $corpus_dir/model/phrase-table -config $corpus_dir/evaluation/$OUTPUT_FILE_NAME.moses.table.ini -lm 0:3:$corpus_dir/evaluation/$OUTPUT_FILE_NAME.moses.table.ini:8`;
|
||||
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $corpus_dir/evaluation/filtered $corpus_dir/evaluation/$OUTPUT_FILE_NAME.moses.table.ini $INPUT_FILE -Binarizer "$MOSES_SRC/bin/processPhraseTable"`;
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $corpus_dir/evaluation/filtered $corpus_dir/evaluation/$OUTPUT_FILE_NAME.moses.table.ini $INPUT_FILE -Binarizer "$MOSES_SRC/bin/CreateOnDiskPt 1 1 4 100 2"`;
|
||||
|
||||
`rm $corpus_dir/evaluation/$OUTPUT_FILE_NAME.moses.table.ini`;
|
||||
|
||||
|
@ -101,7 +101,7 @@ sub run_transliteration
|
||||
|
||||
`$MOSES_SRC/scripts/training/train-model.perl -mgiza -mgiza-cpus 10 -dont-zip -first-step 9 -external-bin-dir $EXTERNAL_BIN_DIR -f $INPUT_EXTENSION -e $OUTPUT_EXTENSION -alignment grow-diag-final-and -parts 5 -reordering msd-bidirectional-fe -score-options '--KneserNey' -phrase-translation-table $TRANSLIT_MODEL/model/phrase-table -reordering-table $TRANSLIT_MODEL/model/reordering-table -config $eval_file.moses.table.ini -lm 0:3:$eval_file.moses.table.ini:8`;
|
||||
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $eval_file.filtered $eval_file.moses.table.ini $eval_file -Binarizer "$MOSES_SRC/bin/processPhraseTable"`;
|
||||
`$MOSES_SRC/scripts/training/filter-model-given-input.pl $eval_file.filtered $eval_file.moses.table.ini $eval_file -Binarizer "$MOSES_SRC/bin/CreateOnDiskPt 1 1 4 100 2"`;
|
||||
|
||||
`rm $eval_file.moses.table.ini`;
|
||||
|
||||
|
@ -180,7 +180,7 @@ sub train_transliteration_module{
|
||||
|
||||
`$MOSES_SRC_DIR/scripts/training/train-model.perl -mgiza -mgiza-cpus 10 -dont-zip -first-step 9 -external-bin-dir $EXTERNAL_BIN_DIR -f $INPUT_EXTENSION -e $OUTPUT_EXTENSION -alignment grow-diag-final-and -parts 5 -score-options '--KneserNey' -phrase-translation-table $OUT_DIR/model/phrase-table -config $OUT_DIR/tuning/moses.table.ini -lm 0:3:$OUT_DIR/tuning/moses.table.ini:8`;
|
||||
|
||||
`$MOSES_SRC_DIR/scripts/training/filter-model-given-input.pl $OUT_DIR/tuning/filtered $OUT_DIR/tuning/moses.table.ini $OUT_DIR/tuning/input -Binarizer "$MOSES_SRC_DIR/bin/processPhraseTable"`;
|
||||
`$MOSES_SRC_DIR/scripts/training/filter-model-given-input.pl $OUT_DIR/tuning/filtered $OUT_DIR/tuning/moses.table.ini $OUT_DIR/tuning/input -Binarizer "$MOSES_SRC_DIR/bin/CreateOnDiskPt 1 1 4 100 2"`;
|
||||
|
||||
`rm $OUT_DIR/tuning/moses.table.ini`;
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user