sshleifer · June 21, 2020 19:06 · sshleifer · Jun 21, 2020
diff --git a/generate_cc25.sh b/generate_cc25.sh

 export langs=ar_AR,cs_CZ,de_DE,en_XX,es_XX,et_EE,fi_FI,fr_XX,gu_IN,hi_IN,it_IT,ja_XX,kk_KZ,ko_KR,lt_LT,lv_LV,my_MM,ne_NP,nl_XX,ro_RO,ru_RU,si_LK,tr_TR,vi_VN,zh_CN
 export CC25=/Users/shleifer/cc25_pretrain
 export outfile=pred_en_ro.txt
 export PRETRAIN=$CC25/model.pt
 fairseq-generate tmp/ --path $PRETRAIN \
 --task translation_from_pretrained_bart -t en_XX -s ro_RO --bpe 'sentencepiece' \
 --sentencepiece-vocab $CC25/sentence.bpe.model --sacrebleu  --remove-bpe 'sentencepiece' \
 --max-sentences 32 --langs $langs --beam 5 > $outfile

 grep ^H $outfile | cut -f3- > tgt.txt
 grep ^T $outfile | cut -f2- > ref.txt

	export langs=ar_AR,cs_CZ,de_DE,en_XX,es_XX,et_EE,fi_FI,fr_XX,gu_IN,hi_IN,it_IT,ja_XX,kk_KZ,ko_KR,lt_LT,lv_LV,my_MM,ne_NP,nl_XX,ro_RO,ru_RU,si_LK,tr_TR,vi_VN,zh_CN
	export CC25=/Users/shleifer/cc25_pretrain
	export outfile=pred_en_ro.txt
	export PRETRAIN=$CC25/model.pt
	fairseq-generate tmp/ --path $PRETRAIN \
	--task translation_from_pretrained_bart -t en_XX -s ro_RO --bpe 'sentencepiece' \
	--sentencepiece-vocab $CC25/sentence.bpe.model --sacrebleu --remove-bpe 'sentencepiece' \
	--max-sentences 32 --langs $langs --beam 5 > $outfile

	grep ^H $outfile \| cut -f3- > tgt.txt
	grep ^T $outfile \| cut -f2- > ref.txt