Skip to content

Commit 3cd81f4

Browse files
committed
Test MidSize
1 parent 56a8a89 commit 3cd81f4

File tree

4 files changed

+58
-2
lines changed

4 files changed

+58
-2
lines changed

scripts/monoTranslationData/Translate.sh

+7-1
Original file line numberDiff line numberDiff line change
@@ -5,6 +5,12 @@ input=$2
55
name=$3
66
translator=$4
77

8+
model=model.pt
9+
if [ $# -ne 4 ]; then
10+
model=$5
11+
fi
12+
13+
814
echo "SET:"$set
915
echo "INPUT:"$input
1016
echo "NAME:"$name
@@ -22,7 +28,7 @@ do
2228
sed -e "s/([0-9]*)//g" $line | sed -e '/\$(.*)/d' | sort -g -k 3,3 | awk '{if($1 != "#") {printf("%s ",$5)}}END{print ""}' | sed -e 's/\,//g' | sed -e 's/\.//g' | sed -e 's/?//g' | sed -e 's/\!//g' | sed -e 's/\"//g' | sed -e 's/^\s*//g' | sed -e 's/\s\s*/ /g' | perl -nle 'print lc' > /tmp/$name/$set.$i.np
2329
/opt/SLT.KIT/scripts/monoTranslationData/ConCat10.pl /tmp/$name/$set.$i.np 10 > /tmp/$name/$set.$i.np.concat
2430
cat /tmp/$name/$set.$i.np.concat | /opt/subword-nmt/apply_bpe.py -c /model/${name}/codec --vocabulary /model/${name}/voc --vocabulary-threshold 50 > /data/${name}/eval/$set.$i.s
25-
/opt/SLT.KIT/scripts/openNMT-py/Translate.sh $set.$i $name $translator
31+
/opt/SLT.KIT/scripts/openNMT-py/Translate.sh $set.$i $name $translator $model
2632
python /opt/SLT.KIT/scripts/monoTranslationData/AddPunctuation.py /tmp/$name/$set.$i.np.concat /data/$translator/eval/$set.$i.t 10 >> /data/${name}/eval/$set.s
2733
((i++))
2834
done < /data/$input/eval/$set/IWSLT.$set/CTM_LIST

scripts/openNMT-py/Translate.sh

+7-1
Original file line numberDiff line numberDiff line change
@@ -4,10 +4,16 @@ set=$1
44
input=$2
55
name=$3
66

7+
model=model.pt
8+
if [ $# -ne 3 ]; then
9+
model=$4
10+
fi
11+
12+
713
mkdir -p /data/$name/eval/
814

915
python -u /opt/OpenNMT-py/translate.py -gpu 0 \
10-
-model /model/$name/model.pt \
16+
-model /model/$name/$model \
1117
-src /data/$input/eval/$set.s \
1218
-batch_size 1 \
1319
-beam_size 16 \

systems/midSize/Download.sh

+7
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,7 @@
1+
#!/bin/bash
2+
3+
export systemName=midSize
4+
rm -r /model/
5+
cd /
6+
wget http://i13pc106.ira.uka.de/~jniehues/SLT.KIT/$sl-$tl/MidSize.tgz
7+
tar -xzvf MidSize.tgz

systems/midSize/Translate.sh

+37
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,37 @@
1+
#!/bin/bash
2+
3+
set=$1
4+
5+
export systemName=midSize
6+
7+
#Download Data - if not there
8+
if [ ! -e /data/orig/eval/$set ]; then
9+
mkdir -p /data/orig/eval/$set
10+
cd /data/orig/eval/$set
11+
wget http://i13pc106.ira.uka.de/~jniehues/IWSLT-SLT/data/eval/$sl-$tl/IWSLT-SLT.$set.$sl-$tl.tgz
12+
tar -xzvf IWSLT-SLT.$set.$sl-$tl.tgz
13+
14+
fi
15+
16+
#Add puncuation
17+
/opt/SLT.KIT/scripts/monoTranslationData/Translate.sh $set orig monoTransPrepro monTrans model.adapt.TED.pt
18+
19+
#Translate
20+
/opt/SLT.KIT/scripts/openNMT-py/Translate.sh $set monoTransPrepro mt model.adapt.TED.pt
21+
22+
#Eval
23+
/opt/SLT.KIT/scripts/evaluate/Eval.sh $set mt
24+
25+
#Prepro manual transcript
26+
/opt/SLT.KIT/scripts/defaultPreprocessor/Translate.sh $set prepro
27+
28+
#Translate manual transcript
29+
/opt/SLT.KIT/scripts/openNMT-py/Translate.sh manualTranscript.$set prepro mt model.adapt.TED.pt
30+
31+
#Eval manual transcript
32+
/opt/SLT.KIT/scripts/evaluate/Eval.manualTranscript.sh $set mt
33+
34+
35+
36+
37+

0 commit comments

Comments
 (0)