Helsinki-NLP/OPUS-MT-train
1
1
Fork 0
mirror of https://github.com/Helsinki-NLP/OPUS-MT-train.git synced 2024-11-27 11:03:13 +03:00
Code Issues Packages Projects Releases Wiki Activity

fixed bug in filtering forward translation recipe

Browse Source
This commit is contained in:
Joerg Tiedemann 2022-03-25 13:53:57 +02:00
parent fc03dd7afe
commit 6f2c9d8bd5
2 changed files with 34 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
lib/generic.mk
View File

@ -285,7 +285,8 @@ FT_SELECTED ?= 95
@for s in ${SRCLANGS}; do \
for t in ${TRGLANGS}; do \
if [ -e ${FORWARDTRANS_HOME}/$$s-$$t/latest ]; then \
if [ ! -e `ls ${FORWARDTRANS_HOME}/$$s-$$t/latest/*.best${FT_SELECTED}.gz | head -1` ]; then \
if [ `ls ${FORWARDTRANS_HOME}/$$s-$$t/latest/ | grep "best${FT_SELECTED}.gz" | wc -l` -eq 0 ]; then \
echo "... extract best translations from $$s-$$t forward translations"; \
${MAKE} -C ${FORWARDTRANS_HOME} SRC=$$s TRG=$$t \
RETAIN=${FT_SELECTED} extract-best-translations; \
fi \

33
lib/projects/elg.mk
View File

@ -86,9 +86,14 @@ elg-ukr-students:
elg-test-tiny2:
${MAKE} EMAIL= SRCLANGS=ukr TRGLANGS=eng test-tiny11-student
${MAKE} EMAIL= SRCLANGS=eng TRGLANGS=ukr test-tiny11-student
${MAKE} EMAIL= SRCLANGS=ukr TRGLANGS=deu STUDENT_DATA=ftbest-bt-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS=eng TRGLANGS=ukr STUDENT_DATA=ftbest-bt-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS=deu TRGLANGS=ukr test-tiny11-student
${MAKE} EMAIL= SRCLANGS=deu TRGLANGS=ukr STUDENT_DATA=ftbest-bt-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS=deu TRGLANGS=ukr STUDENT_DATA=ftbest-ftmono-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS=deu TRGLANGS=ukr STUDENT_DATA=pft-pbt-bt test-tiny11-student
${MAKE} EMAIL= SRCLANGS=ukr TRGLANGS=deu test-tiny11-student
${MAKE} EMAIL= SRCLANGS=ukr TRGLANGS=deu STUDENT_DATA=ftbest-bt-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS=ukr TRGLANGS=deu STUDENT_DATA=ftbest-ftmono-nopar test-tiny11-student
${MAKE} EMAIL= SRCLANGS="ces slk" TRGLANGS=ukr STUDENT_DATA=pft-pbt-bt test-tiny11-student
${MAKE} EMAIL= SRCLANGS=gmq TRGLANGS=ukr STUDENT_DATA=pft-pbt-bt test-tiny11-student
@ -229,6 +234,24 @@ elg-ukr2deu-student3:
elg-fin2ukr-student:
${MAKE} SRCLANGS=fin TRGLANGS=ukr CHECK_TRAINDATA_SIZE=1 CLEAN_CORPUS_TRAINING_DATA=1 STUDENT_DATA=ftbest-ftmono-nopar train-tiny11-student
elg-ukr2fin-student:
${MAKE} SRCLANGS=ukr TRGLANGS=fin CHECK_TRAINDATA_SIZE=1 CLEAN_CORPUS_TRAINING_DATA=1 STUDENT_DATA=ftbest-ftmono-nopar train-tiny11-student
elg-zle2fin-tiny11:
${MAKE} MARIAN_EXTRA=--no-restore-corpus \
DATA_PREPARE_HPCPARAMS='${DATA_PREPARE_HPCPARAMS} CPUJOB_HPC_DISK=1000' \
DATA_ALIGN_HPCPARAMS="${DATA_ALIGN_HPCPARAMS} CPUJOB_HPC_DISK=1000" \
CHECK_TRAINDATA_SIZE=1 CLEAN_CORPUS_TRAINING_DATA=1 \
STUDENT_DATA=ftbest-ftmono-nopar SRCLANGS="ukr rus" TRGLANGS=fin \
LANGPAIRSTR="zle-ukr" train-tiny11-student
elg-spa2ukr-student:
${MAKE} MARIAN_EARLY_STOPPING=15 SRCLANGS=spa TRGLANGS=ukr train-tiny11-student
@ -252,6 +275,9 @@ elg-ukr2eng-student:
elg-eng2ukr-student2:
${MAKE} MARIAN_EARLY_STOPPING=15 STUDENT_DATA=ftbest-bt-nopar SRCLANGS=eng TRGLANGS=ukr train-tiny11-student
elg-ukr2eng-student2:
${MAKE} CONTINUE_EXISTING=1 MARIAN_EARLY_STOPPING=15 SRCLANGS=ukr TRGLANGS=eng STUDENT_DATA=ftbest-ftmono-nopar train-tiny11-student
## missing evaluations and dist packages
@ -282,6 +308,11 @@ elg-dist-missing:
elg-zle2fin-pivot:
${MAKE} MODELTYPE=transformer-big tatoeba-zle2fin-trainjob-pbt-pft-bt
elg-continue-missing:
for l in deu fra ita por spa; do \
${MAKE} MODELTYPE=transformer-big tatoeba-$${l}2zle-trainjob; \