-
Notifications
You must be signed in to change notification settings - Fork 2
/
back_translate_gt_quoted.mk
86 lines (65 loc) · 4.84 KB
/
back_translate_gt_quoted.mk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
####################################################################################################################
##### back translate quoted dataset using GT ###########################################################################
####################################################################################################################
.PRECIOUS: $(foreach name, $(all_names), $(experiment)/oht/%/gt/quoted/$(name).txt)
# with no glossary option we will get output for both glossary and non-glossary translations
# so we use no_glossary only when downloading and post_processing the results
no_glossary ?= --no_glossary
skip_translation ?= false
$(foreach name, $(all_names), $(experiment)/oht/%/gt/quoted/$(name)):
mkdir -p $(experiment)/oht/$*/gt/quoted
# prepare quoted data for back translation to English
for f in $(all_names) ; do python3 $(multilingual_scripts)/text_edit.py --prepare_for_gt --input_file $(experiment)/oht/$*/quoted/$$f.tsv --output_file ./$(experiment)/oht/$*/gt/quoted/$$f.tsv.tmp ; done
for f in $(all_names) ; do cut -f1,2 ./$(experiment)/oht/$*/gt/quoted/$$f.tsv.tmp > ./$(experiment)/oht/$*/gt/quoted/$$f.tsv ; rm -rf ./$(experiment)/oht/$*/gt/quoted/$$f.tsv.tmp ; done
$(experiment)/oht/%/gt/quoted-back: $(foreach name, $(all_names), $(experiment)/oht/%/gt/quoted/$(name))
mkdir -p $@
# upload dataset
python3 $(multilingual_scripts)/translate_v3.py --project_id $(gt_project_id) --project_number $(gt_project_number) --credential_file $(gt_credential_file) \
--update_dataset --input_local_path $(experiment)/oht/$*/gt/quoted --input_names $(foreach name, $(all_names), $(name).tsv) --input_bucket $(input_bucket)
# upload glossary
if ! $(skip_translation) ; then \
python3 $(multilingual_scripts)/translate_v3.py --project_id $(gt_project_id) --project_number $(gt_project_number) --credential_file $(gt_credential_file) \
--update_glossary --glossary_bucket $(glossary_bucket) --input_local_path $(experiment)/oht/$*/gt/quoted --input_names $(foreach name, $(all_names), $(name).tsv) --glossary_type ${glossary_type} --glossary_local_path $(multilingual_scripts)/extras/ --source_lang $* --target_langs en ; \
fi
# batch translate
# you can specify multiple files to be translated simultaneously but we translate each split individually to keep output_buckets separate
if ! $(skip_translation) ; then \
for f in $(all_names) ; do python3 $(multilingual_scripts)/translate_v3.py --project_id $(gt_project_id) --project_number $(gt_project_number) --credential_file $(gt_credential_file) \
--do_translate --input_local_path $(experiment)/oht/$*/gt/quoted --input_names $$f.tsv --output_bucket output_$(experiment)_$*_back_$$f --overwrite_output --glossary_type default \
--input_bucket $(input_bucket) --source_lang $* --target_langs en ; \
done ; \
fi
# download and post_process results
for f in $(all_names) ; do python3 $(multilingual_scripts)/translate_v3.py --project_id $(gt_project_id) --project_number $(gt_project_number) --credential_file $(gt_credential_file) \
--no_glossary --download_results --input_local_path $(experiment)/oht/$*/gt/quoted --input_names $$f.tsv --output_bucket output_$(experiment)_$*_back_$$f --overwrite_output --glossary_type default \
$(no_glossary) --input_bucket $(input_bucket) --source_lang $* --target_langs en --output_local_path $@ ; \
done
$(experiment)/oht/%/gt/quoted-unrefined-back: $(experiment)/oht/%/gt/quoted-back
mkdir -p $@
# insert programs
for f in $(all_names) ; do paste $<$$f.tsv $(experiment)/en/quoted-uniq-progs/$$f.tsv > $@/$$f.tsv ; done
$(experiment)/oht/%/gt/quoted-back-cleaned: $(experiment)/oht/%/gt/quoted-unrefined-back
mkdir -p $@
# fix punctuation and clean dataset
for f in $(all_names) ; do python3 $(multilingual_scripts)/text_edit.py --post_process_translation --refine_sentence --num_columns $(num_columns) --input_file $</$$f.tsv --output_file $@/$$f.tsv ; done
$(experiment)/oht/%/gt/quoted-back-augmented: $(experiment)/oht/%/gt/quoted-back-cleaned
mkdir -p $@
# augment (=unquote) qpis inputs with actual parameter values
for f in $(all_names) ; do \
if ( [ $$f == $(train_name) ] ); then \
$(genie) augment -o $@/$$f.tsv --param-locale en -l en-US $(default_augment_train_hparams) $</$$f.tsv ; \
else \
$(genie) augment -o $@/$$f.tsv --param-locale en -l en-US $(default_augment_evaltest_hparams) $</$$f.tsv ; \
fi ; \
done
$(experiment)/oht/%/gt/final-back: $(experiment)/oht/%/gt/quoted-back-augmented
mkdir -p $@
# remove quotation marks in the sentence
for f in $(all_names) ; do python3 $(multilingual_scripts)/text_edit.py --remove_qpis --input_file $</$$f.tsv --output_file $@/$$f.tsv ; done
back_translate_gt_quoted_%: $(experiment)/oht/%/gt/final-back
# done!
echo $@
back_translate_gt_quoted_all_langs:
for lang in $(gt_oht_languages) ; do \
make -B back_translate_gt_quoted_$$lang ; \
done