s416267/s416267-mlworkshops
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

metryki

Browse Source
This commit is contained in:
Kamil Paprota 2020-04-03 15:02:55 +02:00
parent 5bf1ff8578
commit 965bfc1aae
2 changed files with 12 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Jenkinsfile vendored
View File

@ -11,7 +11,9 @@ pipeline {
sh('chmod +x ./lines.sh')
sh('./lines.sh')
archiveArtifacts 'lines_number.txt'
archiveArtifacts 'WER.txt'
archiveArtifacts 'wer.txt'
archiveArtifacts 'srr.txt'
archiveArtifacts 'wikiniews_results_wer.tsv'
}
}
}

11
lines.sh
View File

@ -1,9 +1,16 @@
wc -l wikiniews_results.tsv > lines_number.txt
cut -f3 wikiniews_results.tsv > reference.txt
cut -f2 wikiniews_results.tsv > hypothesis.txt
cut -f3 wikiniews_results.tsv > reference.txt
awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < reference.txt > reference.trn
awk 'BEGIN{FS=OFS="\t"}{print $0,"(sp1_"NR")"}' < hypothesis.txt > hypothesis.trn
sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout > WER.txt
sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout | grep Scores > scores
awk 'BEGIN{OFS="\t"}{print $6, $7, $8, $9}' < scores | awk '{ print ($2 + $3 + $4)/($2 + $3 + $1) }' > wer.tsv
paste $1 wer.tsv > wikiniews_results_wer.tsv
awk 'BEGIN{FS=OFS="\t"}{ bad = bad + $1; all++ }END{ print bad/all }' < wer.tsv >> wer.txt
awk 'BEGIN{FS=OFS="\t"}{ if ( $1 == 0 ) good++; all++ }END{ print good/all }' < wer.tsv >> srr.txt