s460941/s460941-metrics
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

metrics

Browse Source
This commit is contained in:
s460941 2020-04-23 21:59:51 +02:00
parent 3a9ff47cd7
commit e5bf02b999
1 changed files with 8 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
count.sh
View File

@ -2,15 +2,18 @@ var=$(wc -l "wikiniews_results.tsv")
lines=${var%% *} lines=${var%% *}
echo $lines > "lines_number.txt" echo $lines > "lines_number.txt"
cut -f2 wikiniews_results.tsv > hypothesis.txt cut -f2 wikiniews_results.tsv | awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' > hyp.trn
cut -f3 wikiniews_results.tsv > reference.txt cut -f3 wikiniews_results.tsv | awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' > ref.trn
awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < hypothesis.txt > hypothesis.trn
awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < reference.txt > reference.trn
sclite -f 0 -r ref.trn trn -h hyp.trn trn -e utf-8 -i rm -o pra stdout | grep "Scores:" | sed 's/Scores: (#C #S #D #I) // ' | awk '{ print ($2+$3+$4)/($1+$2+$3)}' > wer_all.txt sclite -f 0 -r ref.trn trn -h hyp.trn trn -e utf-8 -i rm -o pra stdout | grep "Scores:" | sed 's/Scores: (#C #S #D #I) // ' | awk '{ print ($2+$3+$4)/($1+$2+$3)}' > wer_all.txt
awk '{ sum += $1; n++ } END { print sum / n; }' < wer_all.txt >> wer.txt awk '{ sum += $1; n++ } END { print sum / n; }' < wer_all.txt >> wer.txt
#cut -f2 wikiniews_results.tsv > hypothesis.txt
#cut -f3 wikiniews_results.tsv > reference.txt
#awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < hypothesis.txt > hypothesis.trn
#awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < reference.txt > reference.trn
#sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout | grep Scores > Scores #sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout | grep Scores > Scores
#awk 'BEGIN{FS=OFS="\t"}{print $6, $7, $8, $9}' > test.txt #awk 'BEGIN{FS=OFS="\t"}{print $6, $7, $8, $9}' > test.txt