Welcome to mirror list, hosted at ThFree Co, Russian Federation.

setup.sh « multi-source « training « tests - github.com/marian-nmt/marian-regression-tests.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
blob: 31d0814acc39c1ad7ff16d06752edac74f5694fa (plain)
1
2
3
4
5
6
test -f $MRT_DATA/europarl.de-en/corpus.bpe.en || exit 1
test -f $MRT_DATA/europarl.de-en/corpus.bpe.de || exit 1

test -f train.bpe.en || head -n 10000 $MRT_DATA/europarl.de-en/corpus.bpe.en > train.bpe.en
test -f train.bpe.de || head -n 10000 $MRT_DATA/europarl.de-en/corpus.bpe.de > train.bpe.de
test -f train.bpe.xx || sed -e 's/\([^ ]\{,3\}\)[^ ]*/\1/g' -e 's/[.,:;?!()]\s\?//g' train.bpe.en > train.bpe.xx