-
Notifications
You must be signed in to change notification settings - Fork 162
/
test.sh
executable file
·752 lines (673 loc) · 36.6 KB
/
test.sh
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
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
#!/usr/bin/env bash
# Copyright 2017 Amazon.com, Inc. or its affiliates. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License"). You may not
# use this file except in compliance with the License. A copy of the License
# is located at
#
# http://aws.amazon.com/apache2.0/
#
# or in the "license" file accompanying this file. This file is distributed on
# an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either
# express or implied. See the License for the specific language governing
# permissions and limitations under the License.
# Confirms that BLEU scores computed by sacreBLEU are the same as Moses' mteval-v13a.pl.
# Note that this doesn't work if you lowercase the data (remove -c to mteval-v13a.pl,
# or add "-lc" to sacreBLEU) because mteval-v13a.pl does not lowercase properly: it uses
# tr/[A-Z]/[a-z], which doesn't cover uppercase letters with diacritics. Also, the
# Chinese preprocessing was applied to all zh sources, references, and system outputs
# (http://statmt.org/wmt17/tokenizeChinese.py), as was done for WMT17.
set -u
if [[ $(echo $BASH_VERSION | cut -d. -f1) -lt 4 ]]; then
echo "This script requires BASH version 4 or above (since it uses hashes)."
exit 1
fi
# Switch from JSON output to text
export SACREBLEU_FORMAT="text"
# Cleanup temporary files
trap "rm -f .tmp* data/.tmp*" EXIT INT TERM
# For Travis CI to work on Windows/Mac OS X
if [[ "$OSTYPE" == "linux-gnu"* ]]; then
PYTHON="python"
CMD="$PYTHON -m sacrebleu"
elif [[ "$OSTYPE" == "msys" ]]; then
PYTHON="python"
CMD="$PYTHON -m sacrebleu"
elif [[ "$OSTYPE" == "darwin"* ]]; then
# OS X ships python -> python2 by default, be explicit
PYTHON=python3
CMD="$PYTHON -m sacrebleu"
fi
export SACREBLEU=$(pwd)/.sacrebleu
export PYTHONPATH="${PWD}" # assuming PYTHONPATH=. as the default
export NO_COLOR=1
# Only run this test
limit_test=${1:-}
SKIP_INITIAL=${SKIP_INITIAL:-}
SKIP_CHRF=${SKIP_CHRF:-}
SKIP_TER=${SKIP_TER:-}
SKIP_MECAB=${SKIP_MECAB:-}
SKIP_MECAB_KO=${SKIP_MECAB_KO:-}
SKIP_MTEVAL13=${SKIP_MTEVAL13:-}
SKIP_MTEVAL14=${SKIP_MTEVAL14:-}
# test case counter
declare -i i=0
# TEST 1: download and process WMT17 data
[[ -d $SACREBLEU/wmt17 ]] && rm -f $SACREBLEU/wmt17/{en-*,*-en*}
${CMD} --echo src -t wmt17 -l cs-en > /dev/null
# Test concatenation of multiple test sets, --echo, -w, --origlang, --verbose
# and a pipeline with two sacrebleu processes
declare -A EXPECTED
EXPECTED["${CMD} -t wmt16,wmt17 -l en-fi --echo ref | ${CMD} -b -w 4 -t wmt16/B,wmt17/B -l en-fi"]=53.7432
EXPECTED["${CMD} -t wmt16,wmt17 -l en-fi --echo ref | ${CMD} -b -w 4 -t wmt16/B,wmt17/B -l en-fi --origlang=en"]=18.9054
EXPECTED["${CMD} -t wmt17 -l en-fi --echo ref | ${CMD} -b -t wmt17/B -l en-fi --detail"]="55.6
origlang=en domain=ALL : sentences=1502 BLEU = 21.4
origlang=fi domain=ALL : sentences=1500 BLEU = 100.0"
EXPECTED["${CMD} -t wmt18,wmt19 -l en-de --echo=src | ${CMD} -t wmt18,wmt19 -l en-de -b --detail"]="3.6
origlang=de domain=ALL : sentences=1498 BLEU = 3.6
origlang=en domain=ALL : sentences=3497 BLEU = 3.5
origlang=en domain=business : sentences=241 BLEU = 3.4
origlang=en country=EU : sentences=265 BLEU = 2.5
origlang=en country=GB : sentences=913 BLEU = 3.1
origlang=en country=OTHER : sentences=801 BLEU = 2.5
origlang=en country=US : sentences=1518 BLEU = 4.2
origlang=en domain=crime : sentences=570 BLEU = 3.6
origlang=en domain=entertainment : sentences=322 BLEU = 5.1
origlang=en domain=politics : sentences=959 BLEU = 3.0
origlang=en domain=scitech : sentences=211 BLEU = 3.1
origlang=en domain=sport : sentences=534 BLEU = 3.6
origlang=en domain=world : sentences=660 BLEU = 3.1"
[[ ! -d data ]] && mkdir data
cd data
if [[ ! -d wmt17-submitted-data ]]; then
echo "Downloading and unpacking WMT'17 system submissions (46 MB)..."
wget -q https://data.statmt.org/wmt17/translation-task/wmt17-submitted-data-v1.0.tgz
tar xzf wmt17-submitted-data-v1.0.tgz
fi
if [[ ! -d en-ja-translation-example-master ]]; then
echo "Downloading and unpacking English-Japanese test data..."
wget -q https://github.com/MorinoseiMorizo/en-ja-translation-example/archive/master.zip
unzip master.zip
rm master.zip
fi
if [[ ! -d en-ko-translation-example-master ]]; then
echo "Downloading and unpacking English-Korean test data..."
wget -q https://github.com/NoUnique/en-ko-translation-example/archive/master.zip
unzip master.zip
rm master.zip
fi
if [ -z $SKIP_INITIAL ]; then
for command in "${!EXPECTED[@]}"; do
echo Testing $command
# Convert line endings to UNIX for Windows tests
obtained=`eval $command | tr -d '\015'`
expected=${EXPECTED[$command]}
if [[ $obtained != $expected ]]; then
echo -e "\nFAILED:\n expected = $expected\n obtained = $obtained"
exit 1
fi
echo PASS
done
# Test loading via file instead of STDIN
echo "Testing loading via file instead of STDIN"
${CMD} -t wmt17 -l en-de --echo ref > .tmp.wmt17
score=$(${CMD} -t wmt17 -l en-de -i .tmp.wmt17 -b)
if [[ $score != '100.0' ]]; then
echo "File test failed."
exit 1
fi
echo PASS
# Test echoing of source, reference, and both
# Replace \r\n with \n for Windows compatibility
${CMD} -t wmt17/ms -l zh-en --echo src | tr -d '\015' > .tmp.echo
diff .tmp.echo $SACREBLEU/wmt17/ms/wmt17_ms.zh-en.src
if [[ $? -ne 0 ]]; then
echo "Source echo failed."
exit 1
fi
${CMD} -t wmt17/ms -l zh-en --echo ref:2 | tr -d '\015' > .tmp.echo
diff .tmp.echo $SACREBLEU/wmt17/ms/wmt17_ms.zh-en.ref-2 > /dev/null
if [[ $? -ne 0 ]]; then
echo "Source echo failed."
pwd
exit 1
fi
export LC_ALL=C
echo '-----------------------'
echo 'Control character tests'
echo '-----------------------'
printf "Hello! How are you \r doing today?" > .tmp_ref_buggy
echo "Hello! How are you doing today?" > .tmp_ref_okay
score1=$( echo "Hello! How are you doing today?" | ${CMD} -w 2 -b .tmp_ref_buggy )
score2=$( echo "Hello! How are you doing today?" | ${CMD} -w 2 -b .tmp_ref_okay )
if [[ $score1 != $score2 ]]; then
echo "Control character in reference test failed"
exit 1
fi
let i++
echo "Passed control character in reference test"
#####################################################################
# Tests for single-ref BLEU, multi-ref BLEU, signature and tokenizers
#####################################################################
path="wmt17-submitted-data/txt/system-outputs/newstest2017/cs-en"
ref1="${path}/newstest2017.online-A.0.cs-en"
ref2="${path}/newstest2017.online-B.0.cs-en"
sys="${path}/newstest2017.PJATK.4760.cs-en"
echo '--------------------------'
echo 'BLEU/CHRF regression tests'
echo '--------------------------'
unset EXPECTED
declare -A EXPECTED
paste $ref1 $ref2 > .tmp_refs
# Single ref, tokenizer variants, lowercase
EXPECTED["${CMD} -w 4 -b -l cs-en $ref1 -i $sys"]=36.8799
EXPECTED["${CMD} -lc -w 4 -b -l cs-en $ref1 -i $sys"]=38.1492
EXPECTED["${CMD} --tokenize 13a -w 4 -b -l cs-en $ref1 -i $sys"]=36.8799
EXPECTED["${CMD} --tokenize none -w 4 -b -l cs-en $ref1 -i $sys"]=34.0638
# multiple REF files
EXPECTED["${CMD} -w 4 -b -l cs-en $ref1 $ref2 -i $sys"]=44.6732
# multiple REF CHRF (epsilon smoothing)
EXPECTED["${CMD} -m chrf --chrf-eps-smoothing -w 4 -b -l cs-en $ref1 $ref2 -i $sys"]=67.8596
# multiple REF CHRF (effective order smoothing)
EXPECTED["${CMD} -m chrf -w 4 -b -l cs-en $ref1 $ref2 -i $sys"]=67.8603
# multiple REF CHRF++
EXPECTED["${CMD} -m chrf -cw 2 -w 4 -b -l cs-en $ref1 $ref2 -i $sys"]=66.1016
# multiple REFs with tab-delimited stream
EXPECTED["${CMD} -w 4 -b -l cs-en --num-refs 2 .tmp_refs -i $sys"]=44.6732
# Check signature correctness for multi-reference
# separate files
EXPECTED["${CMD} -l cs-en $ref1 $ref2 -i $sys | perl -pe 's/.*nrefs:([0-9]).*/\1/'"]=2
# tab delimited stream
EXPECTED["${CMD} -l cs-en --num-refs 2 .tmp_refs -i $sys | perl -pe 's/.*nrefs:([0-9]).*/\1/'"]=2
# Run the tests
for command in "${!EXPECTED[@]}"; do
echo Testing $command
obtained=`eval $command`
expected=${EXPECTED[$command]}
if [[ $obtained != $expected ]]; then
echo -e "\nFAILED:\n expected = $expected\n obtained = $obtained"
exit 1
fi
echo PASS
let i++
done
fi
#######################################################
# Pre-computed chrF scores from official implementation
# Cmd: chrF++.py -H hyp -R ref -nw 0
#######################################################
declare -A CHRF=( ["newstest2017.PJATK.4760.cs-en.sgm"]=52.5947
["newstest2017.online-A.0.cs-en.sgm"]=53.3856
["newstest2017.CU-Chimera.4886.en-cs.sgm"]=48.3370
["newstest2017.LIUM-FNMT.4852.en-cs.sgm"]=48.4708
["newstest2017.C-3MA.4959.en-de.sgm"]=51.9533
["newstest2017.FBK.4870.en-de.sgm"]=54.7152
["newstest2017.AaltoHnmtFlatcat.4798.en-fi.sgm"]=50.5981
["newstest2017.online-A.0.en-fi.sgm"]=49.5458
["newstest2017.limsi-factored-norm.5041.en-lv.sgm"]=49.3634
["newstest2017.online-A.0.en-lv.sgm"]=45.2101
["newstest2017.jhu-pbmt.4986.en-ru.sgm"]=54.9569
["newstest2017.online-B.0.en-ru.sgm"]=60.4059
["newstest2017.LIUM-NMT.4953.en-tr.sgm"]=47.5881
["newstest2017.online-G.0.en-tr.sgm"]=48.7404
["newstest2017.UU-HNMT.5134.en-zh.sgm"]=22.6844
["newstest2017.online-F.0.en-zh.sgm"]=18.7403
)
#########################################################
# Pre-computed chrF++ scores from official implementation
# Cmd: chrF++.py -H hyp -R ref
#########################################################
declare -A CHRFPP=( ["newstest2017.PJATK.4760.cs-en.sgm"]=50.2947
["newstest2017.online-A.0.cs-en.sgm"]=51.1037
["newstest2017.CU-Chimera.4886.en-cs.sgm"]=45.6732
["newstest2017.LIUM-FNMT.4852.en-cs.sgm"]=45.7210
["newstest2017.C-3MA.4959.en-de.sgm"]=49.1683
["newstest2017.FBK.4870.en-de.sgm"]=52.2330
["newstest2017.AaltoHnmtFlatcat.4798.en-fi.sgm"]=46.6295
["newstest2017.online-A.0.en-fi.sgm"]=44.9147
["newstest2017.limsi-factored-norm.5041.en-lv.sgm"]=45.7117
["newstest2017.online-A.0.en-lv.sgm"]=40.8963
["newstest2017.jhu-pbmt.4986.en-ru.sgm"]=52.0154
["newstest2017.online-B.0.en-ru.sgm"]=57.6404
["newstest2017.LIUM-NMT.4953.en-tr.sgm"]=43.8311
["newstest2017.online-G.0.en-tr.sgm"]=44.6236
["newstest2017.UU-HNMT.5134.en-zh.sgm"]=17.0181
["newstest2017.online-F.0.en-zh.sgm"]=14.1572
)
if [ -z $SKIP_CHRF ]; then
echo "------------------------------"
echo "Starting chrF and chrF++ tests"
echo "------------------------------"
# Test only for different target languages as there is no tokenization
# issue involved in chrF
for pair in cs-en en-cs en-de en-fi en-lv en-ru en-tr en-zh; do
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for sgm in wmt17-submitted-data/sgm/system-outputs/newstest2017/$pair/*.sgm; do
name=$(basename $sgm)
if [[ ! -v CHRF[$name] ]]; then continue; fi
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
sys=$(basename $sgm .sgm | perl -pe 's/newstest2017\.//')
txt=$(dirname $sgm | perl -pe 's/sgm/txt/')/$(basename $sgm .sgm)
# Test chrF
score=$(cat $txt | ${CMD} -w 4 -t wmt17 -l $source-$target -b --metrics chrf --chrf-eps-smoothing)
expected_score=${CHRF[$name]}
echo "import sys; sys.exit(1 if abs(${score}-${expected_score}) > 1e-6 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED chrF test $pair/$sys (wanted $expected_score got $score)"
exit 1
fi
echo "Passed $source-$target $sys chrF++.py: $expected_score sacreCHRF: $score"
let i++
# Test chrF++
score=$(cat $txt | ${CMD} -w 4 -t wmt17 -l $source-$target -b --metrics chrf --chrf-word-order 2 --chrf-eps-smoothing)
expected_score=${CHRFPP[$name]}
echo "import sys; sys.exit(1 if abs(${score}-${expected_score}) > 1e-6 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED chrF++ test $pair/$sys (wanted $expected_score got $score)"
exit 1
fi
echo "Passed $source-$target $sys chrF++.py: $expected_score sacreCHRF++: $score"
let i++
done
done
fi
################################################################
# Pre-computed results from Moses' mteval-v13a.pl for BLEU tests
################################################################
echo "------------------------------------"
echo "Starting BLEU tests (mteval-v13a.pl)"
echo "------------------------------------"
declare -A MTEVAL=( ["newstest2017.PJATK.4760.cs-en.sgm"]=23.15
["newstest2017.online-A.0.cs-en.sgm"]=25.12
["newstest2017.online-B.0.cs-en.sgm"]=27.45
["newstest2017.uedin-nmt.4955.cs-en.sgm"]=30.95
["newstest2017.C-3MA.4958.de-en.sgm"]=28.98
["newstest2017.KIT.4951.de-en.sgm"]=34.56
["newstest2017.LIUM-NMT.4733.de-en.sgm"]=30.1
["newstest2017.RWTH-nmt-ensemble.4920.de-en.sgm"]=33.12
["newstest2017.SYSTRAN.4846.de-en.sgm"]=33.17
["newstest2017.TALP-UPC.4830.de-en.sgm"]=28.09
["newstest2017.online-A.0.de-en.sgm"]=32.48
["newstest2017.online-B.0.de-en.sgm"]=32.97
["newstest2017.online-F.0.de-en.sgm"]=17.67
["newstest2017.online-G.0.de-en.sgm"]=26.04
["newstest2017.uedin-nmt.4723.de-en.sgm"]=35.12
["newstest2017.CU-Chimera.4886.en-cs.sgm"]=20.51
["newstest2017.LIUM-FNMT.4852.en-cs.sgm"]=19.95
["newstest2017.LIUM-NMT.4947.en-cs.sgm"]=20.2
["newstest2017.PJATK.4761.en-cs.sgm"]=16.18
["newstest2017.limsi-factored-norm.4957.en-cs.sgm"]=20.22
["newstest2017.online-A.0.en-cs.sgm"]=16.55
["newstest2017.online-B.0.en-cs.sgm"]=20.12
["newstest2017.tuning-task-afrl_4gb.sgm.0.en-cs.sgm"]=14.18
["newstest2017.tuning-task-afrl_8gb.sgm.0.en-cs.sgm"]=14.69
["newstest2017.tuning-task-baseline_4gb.sgm.0.en-cs.sgm"]=13.69
["newstest2017.tuning-task-baseline_8gb.sgm.0.en-cs.sgm"]=13.79
["newstest2017.tuning-task-denisov_4gb.sgm.0.en-cs.sgm"]=12.64
["newstest2017.tuning-task-ufal_4gb.sgm.0.en-cs.sgm"]=13.06
["newstest2017.tuning-task-ufal_8gb.sgm.0.en-cs.sgm"]=15.25
["newstest2017.uedin-nmt.4956.en-cs.sgm"]=22.8
["newstest2017.C-3MA.4959.en-de.sgm"]=22.67
["newstest2017.FBK.4870.en-de.sgm"]=26.33
["newstest2017.KIT.4950.en-de.sgm"]=26.08
["newstest2017.LIUM-NMT.4900.en-de.sgm"]=26.6
["newstest2017.LMU-nmt-reranked.4934.en-de.sgm"]=27.11
["newstest2017.LMU-nmt-single.4893.en-de.sgm"]=26.56
["newstest2017.PROMT-Rule-based.4735.en-de.sgm"]=16.61
["newstest2017.RWTH-nmt-ensemble.4921.en-de.sgm"]=26.02
["newstest2017.SYSTRAN.4847.en-de.sgm"]=26.71
["newstest2017.TALP-UPC.4834.en-de.sgm"]=21.24
["newstest2017.online-A.0.en-de.sgm"]=20.8
["newstest2017.online-B.0.en-de.sgm"]=26.65
["newstest2017.online-F.0.en-de.sgm"]=15.45
["newstest2017.online-G.0.en-de.sgm"]=18.16
["newstest2017.uedin-nmt.4722.en-de.sgm"]=28.3
["newstest2017.xmu.4910.en-de.sgm"]=26.69
["newstest2017.AaltoHnmtFlatcat.4798.en-fi.sgm"]=17.15
["newstest2017.AaltoHnmtMultitask.4873.en-fi.sgm"]=20.28
["newstest2017.HY-AH.4797.en-fi.sgm"]=9.33
["newstest2017.HY-HNMT.4961.en-fi.sgm"]=20.72
["newstest2017.HY-SMT.4882.en-fi.sgm"]=15.87
["newstest2017.TALP-UPC.4939.en-fi.sgm"]=11.47
["newstest2017.apertium-unconstrained.4769.en-fi.sgm"]=1.05
["newstest2017.jhu-nmt-lattice-rescore.4903.en-fi.sgm"]=15.96
["newstest2017.jhu-pbmt.4968.en-fi.sgm"]=14.46
["newstest2017.online-A.0.en-fi.sgm"]=13.72
["newstest2017.online-B.0.en-fi.sgm"]=22.04
["newstest2017.online-G.0.en-fi.sgm"]=15.51
["newstest2017.C-3MA.5069.en-lv.sgm"]=13.64
["newstest2017.HY-HNMT.5066.en-lv.sgm"]=16.76
["newstest2017.KIT.5062.en-lv.sgm"]=18.31
["newstest2017.LIUM-FNMT.5043.en-lv.sgm"]=16.19
["newstest2017.LIUM-NMT.5042.en-lv.sgm"]=17.01
["newstest2017.PJATK.4744.en-lv.sgm"]=10.15
["newstest2017.QT21-System-Combination.5063.en-lv.sgm"]=18.63
["newstest2017.jhu-pbmt.4969.en-lv.sgm"]=14.35
["newstest2017.limsi-factored-norm.5041.en-lv.sgm"]=16.63
["newstest2017.online-A.0.en-lv.sgm"]=10.84
["newstest2017.online-B.0.en-lv.sgm"]=17.91
["newstest2017.tilde-c-nmt-smt-hybrid.5049.en-lv.sgm"]=20.14
["newstest2017.tilde-nc-nmt-smt-hybrid.5047.en-lv.sgm"]=20.79
["newstest2017.tilde-nc-smt.5044.en-lv.sgm"]=21.05
["newstest2017.uedin-nmt.5016.en-lv.sgm"]=16.93
["newstest2017.usfd-consensus-kit.5078.en-lv.sgm"]=17.48
["newstest2017.usfd-consensus-qt21.5077.en-lv.sgm"]=17.99
["newstest2017.PROMT-Rule-based.4736.en-ru.sgm"]=22.3
["newstest2017.afrl-mitll-backtrans.4907.en-ru.sgm"]=25.37
["newstest2017.jhu-pbmt.4986.en-ru.sgm"]=25.32
["newstest2017.online-A.0.en-ru.sgm"]=23.83
["newstest2017.online-B.0.en-ru.sgm"]=32.05
["newstest2017.online-F.0.en-ru.sgm"]=10.08
["newstest2017.online-G.0.en-ru.sgm"]=27.11
["newstest2017.online-H.0.en-ru.sgm"]=28.41
["newstest2017.uedin-nmt.4756.en-ru.sgm"]=29.79
["newstest2017.JAIST.4858.en-tr.sgm"]=9.98
["newstest2017.LIUM-NMT.4953.en-tr.sgm"]=16.03
["newstest2017.jhu-nmt-lattice-rescore.4904.en-tr.sgm"]=10.37
["newstest2017.jhu-pbmt.4970.en-tr.sgm"]=9.81
["newstest2017.online-A.0.en-tr.sgm"]=11.62
["newstest2017.online-B.0.en-tr.sgm"]=19.93
["newstest2017.online-G.0.en-tr.sgm"]=13.93
["newstest2017.uedin-nmt.4932.en-tr.sgm"]=16.43
["newstest2017.CASICT-DCU-NMT.5157.en-zh.sgm"]=30.52
["newstest2017.Oregon-State-University-S.5174.en-zh.sgm"]=25.93
["newstest2017.SogouKnowing-nmt.5131.en-zh.sgm"]=34.87
["newstest2017.UU-HNMT.5134.en-zh.sgm"]=23.92
["newstest2017.jhu-nmt.5153.en-zh.sgm"]=31.05
["newstest2017.online-A.0.en-zh.sgm"]=29.02
["newstest2017.online-B.0.en-zh.sgm"]=32.07
["newstest2017.online-F.0.en-zh.sgm"]=18.58
["newstest2017.online-G.0.en-zh.sgm"]=21.31
["newstest2017.uedin-nmt.5111.en-zh.sgm"]=36.28
["newstest2017.xmunmt.5165.en-zh.sgm"]=35.84
["newstest2017.Hunter-MT.4925.fi-en.sgm"]=19.96
["newstest2017.TALP-UPC.4937.fi-en.sgm"]=16.13
["newstest2017.apertium-unconstrained.4793.fi-en.sgm"]=6.13
["newstest2017.online-A.0.fi-en.sgm"]=21.13
["newstest2017.online-B.0.fi-en.sgm"]=27.62
["newstest2017.online-G.0.fi-en.sgm"]=22.48
["newstest2017.C-3MA.5067.lv-en.sgm"]=14.31
["newstest2017.Hunter-MT.5092.lv-en.sgm"]=16.23
["newstest2017.PJATK.4740.lv-en.sgm"]=12.42
["newstest2017.jhu-pbmt.4980.lv-en.sgm"]=16.79
["newstest2017.online-A.0.lv-en.sgm"]=15.72
["newstest2017.online-B.0.lv-en.sgm"]=22.02
["newstest2017.tilde-c-nmt-smt-hybrid.5051.lv-en.sgm"]=20.0
["newstest2017.tilde-nc-nmt-smt-hybrid.5050.lv-en.sgm"]=21.92
["newstest2017.uedin-nmt.5017.lv-en.sgm"]=18.98
["newstest2017.NRC.4855.ru-en.sgm"]=33.88
["newstest2017.afrl-mitll-opennmt.4896.ru-en.sgm"]=33.95
["newstest2017.afrl-mitll-syscomb.4905.ru-en.sgm"]=34.71
["newstest2017.jhu-pbmt.4978.ru-en.sgm"]=31.47
["newstest2017.online-A.0.ru-en.sgm"]=31.18
["newstest2017.online-B.0.ru-en.sgm"]=36.94
["newstest2017.online-F.0.ru-en.sgm"]=15.9
["newstest2017.online-G.0.ru-en.sgm"]=34.51
["newstest2017.uedin-nmt.4890.ru-en.sgm"]=30.77
["newstest2017.JAIST.4859.tr-en.sgm"]=12.42
["newstest2017.LIUM-NMT.4888.tr-en.sgm"]=17.91
["newstest2017.PROMT-SMT.4737.tr-en.sgm"]=13.52
["newstest2017.afrl-mitll-m2w-nr1.4901.tr-en.sgm"]=17.54
["newstest2017.afrl-mitll-syscomb.4902.tr-en.sgm"]=18.05
["newstest2017.jhu-pbmt.4972.tr-en.sgm"]=12.64
["newstest2017.online-A.0.tr-en.sgm"]=22.18
["newstest2017.online-B.0.tr-en.sgm"]=25.62
["newstest2017.online-G.0.tr-en.sgm"]=16.38
["newstest2017.uedin-nmt.4931.tr-en.sgm"]=20.08
["newstest2017.CASICT-DCU-NMT.5144.zh-en.sgm"]=22.32
["newstest2017.NMT-Model-Average-Multi-Cards.5099.zh-en.sgm"]=18.98
["newstest2017.NRC.5172.zh-en.sgm"]=25.78
["newstest2017.Oregon-State-University-S.5173.zh-en.sgm"]=18.51
["newstest2017.PROMT-SMT.5125.zh-en.sgm"]=16.47
["newstest2017.ROCMT.5183.zh-en.sgm"]=20.76
["newstest2017.SogouKnowing-nmt.5171.zh-en.sgm"]=26.38
["newstest2017.UU-HNMT.5162.zh-en.sgm"]=15.94
["newstest2017.afrl-mitll-opennmt.5109.zh-en.sgm"]=21.28
["newstest2017.jhu-nmt.5151.zh-en.sgm"]=20.48
["newstest2017.online-A.0.zh-en.sgm"]=24.51
["newstest2017.online-B.0.zh-en.sgm"]=33.23
["newstest2017.online-F.0.zh-en.sgm"]=11.39
["newstest2017.online-G.0.zh-en.sgm"]=15.5
["newstest2017.uedin-nmt.5112.zh-en.sgm"]=25.7
["newstest2017.xmunmt.5160.zh-en.sgm"]=26.0
["kyoto-test"]=14.48
["flores101.devtest.en-ko"]=33.18
)
if [ -z $SKIP_MTEVAL13 ]; then
for pair in cs-en de-en en-cs en-de en-fi en-lv en-ru en-tr en-zh fi-en lv-en ru-en tr-en zh-en; do
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for sgm in wmt17-submitted-data/sgm/system-outputs/newstest2017/$pair/*.sgm; do
name=$(basename $sgm)
if [[ ! -v MTEVAL[$name] ]]; then continue; fi
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
sys=$(basename $sgm .sgm | perl -pe 's/newstest2017\.//')
txt=$(dirname $sgm | perl -pe 's/sgm/txt/')/$(basename $sgm .sgm)
src=wmt17-submitted-data/sgm/sources/newstest2017-$source$target-src.$source.sgm
ref=wmt17-submitted-data/sgm/references/newstest2017-$source$target-ref.$target.sgm
# mteval=$($MOSES/scripts/generic/mteval-v13a.pl -c -s $src -r $ref -t $sgm 2> /dev/null | grep "BLEU score" | cut -d' ' -f9)
# mteval=$(echo "print($bleu1 * 100)" | python)
score=$(cat $txt | ${CMD} -w 2 -t wmt17 -l $source-$target -b)
echo "import sys; sys.exit(1 if abs($score-${MTEVAL[$name]}) > 0.01 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED test $pair/$sys (wanted ${MTEVAL[$name]} got $score)"
exit 1
fi
echo "Passed $source-$target $sys mteval-v13a.pl: ${MTEVAL[$name]} sacreBLEU: $score"
let i++
done
done
fi
############################################################################
# Pre-computed results from Moses' mteval-v14.pl for BLEU tests
# mteval-v14a.pl -c -s $src -r $ref -t $sgm -b --international-tokenization
############################################################################
echo "-----------------------------------------------------------------"
echo "Starting BLEU tests (mteval-v14.pl, --international-tokenization)"
echo "-----------------------------------------------------------------"
declare -A MTEVAL14=( ["newstest2017.online-A.0.en-ru.sgm"]=23.99
["newstest2017.online-A.0.en-cs.sgm"]=16.65
["newstest2017.uedin-nmt.4890.ru-en.sgm"]=30.91
["newstest2017.xmunmt.5160.zh-en.sgm"]=26.8
["newstest2017.LIUM-NMT.4900.en-de.sgm"]=26.89
["newstest2017.tilde-nc-nmt-smt-hybrid.5047.en-lv.sgm"]=20.88
["newstest2017.TALP-UPC.4937.fi-en.sgm"]=16.18
["newstest2017.LIUM-NMT.4888.tr-en.sgm"]=17.94
["newstest2017.SogouKnowing-nmt.5131.en-zh.sgm"]=7.24
["newstest2017.online-B.0.en-tr.sgm"]=22.88
["newstest2017.apertium-unconstrained.4769.en-fi.sgm"]=1.08
["newstest2017.tilde-c-nmt-smt-hybrid.5051.lv-en.sgm"]=20.44
["newstest2017.RWTH-nmt-ensemble.4920.de-en.sgm"]=33.73
# Below two not exactly compatible with mteval due to 's (#138)
["newstest2017.PJATK.4760.cs-en.sgm"]=23.47
["newstest2017.PJATK.4761.en-cs.sgm"]=15.91
)
if [ -z $SKIP_MTEVAL14 ]; then
for pair in cs-en de-en en-cs en-de en-fi en-lv en-ru en-tr en-zh fi-en lv-en ru-en tr-en zh-en; do
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for sgm in wmt17-submitted-data/sgm/system-outputs/newstest2017/$pair/*.sgm; do
name=$(basename $sgm)
if [[ ! -v MTEVAL14[$name] ]]; then continue; fi
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
sys=$(basename $sgm .sgm | perl -pe 's/newstest2017\.//')
txt=$(dirname $sgm | perl -pe 's/sgm/txt/')/$(basename $sgm .sgm)
src=wmt17-submitted-data/sgm/sources/newstest2017-$source$target-src.$source.sgm
ref=wmt17-submitted-data/sgm/references/newstest2017-$source$target-ref.$target.sgm
score=$(cat $txt | ${CMD} -w 2 -t wmt17 -l $source-$target -b --tokenize intl)
echo "import sys; sys.exit(1 if abs($score-${MTEVAL14[$name]}) > 0.01 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED test $pair/$sys (wanted ${MTEVAL14[$name]} got $score)"
exit 1
fi
echo "Passed $source-$target $sys mteval-v14.pl: ${MTEVAL14[$name]} sacreBLEU: $score"
let i++
done
done
fi
#######################################################
# Pre-computed TER scores from official implementation
# Cmd: java -jar tercom.7.25.jar -r REF -h HYP | grep '^Total TER' | awk '{ printf "%.3f\n", $3 }'
#######################################################
declare -A TER=( ["newstest2017.online-A.0.cs-en.sgm"]=0.637
["newstest2017.online-B.0.cs-en.sgm"]=0.605
["newstest2017.PJATK.4760.cs-en.sgm"]=0.664
["newstest2017.uedin-nmt.4955.cs-en.sgm"]=0.584
["newstest2017.CU-Chimera.4886.en-cs.sgm"]=0.696
["newstest2017.limsi-factored-norm.4957.en-cs.sgm"]=0.696
["newstest2017.LIUM-FNMT.4852.en-cs.sgm"]=0.699
["newstest2017.LIUM-NMT.4947.en-cs.sgm"]=0.701
["newstest2017.online-A.0.en-cs.sgm"]=0.743
["newstest2017.online-B.0.en-cs.sgm"]=0.703
["newstest2017.PJATK.4761.en-cs.sgm"]=0.757
["newstest2017.uedin-nmt.4956.en-cs.sgm"]=0.667
["newstest2017.C-3MA.4959.en-de.sgm"]=0.669
["newstest2017.FBK.4870.en-de.sgm"]=0.636
["newstest2017.KIT.4950.en-de.sgm"]=0.641
["newstest2017.LIUM-NMT.4900.en-de.sgm"]=0.633
["newstest2017.LMU-nmt-reranked.4934.en-de.sgm"]=0.618
["newstest2017.LMU-nmt-single.4893.en-de.sgm"]=0.626
["newstest2017.online-A.0.en-de.sgm"]=0.690
["newstest2017.online-B.0.en-de.sgm"]=0.632
["newstest2017.online-F.0.en-de.sgm"]=0.756
["newstest2017.online-G.0.en-de.sgm"]=0.733
["newstest2017.PROMT-Rule-based.4735.en-de.sgm"]=0.752
["newstest2017.RWTH-nmt-ensemble.4921.en-de.sgm"]=0.628
["newstest2017.SYSTRAN.4847.en-de.sgm"]=0.611
["newstest2017.TALP-UPC.4834.en-de.sgm"]=0.685
["newstest2017.uedin-nmt.4722.en-de.sgm"]=0.612
["newstest2017.xmu.4910.en-de.sgm"]=0.622
["newstest2017.AaltoHnmtFlatcat.4798.en-fi.sgm"]=0.750
["newstest2017.AaltoHnmtMultitask.4873.en-fi.sgm"]=0.673
["newstest2017.apertium-unconstrained.4769.en-fi.sgm"]=1.246
["newstest2017.HY-AH.4797.en-fi.sgm"]=0.857
["newstest2017.HY-HNMT.4961.en-fi.sgm"]=0.676
["newstest2017.HY-SMT.4882.en-fi.sgm"]=0.754
["newstest2017.jhu-nmt-lattice-rescore.4903.en-fi.sgm"]=0.758
["newstest2017.jhu-pbmt.4968.en-fi.sgm"]=0.778
["newstest2017.online-A.0.en-fi.sgm"]=0.761
["newstest2017.online-B.0.en-fi.sgm"]=0.665
["newstest2017.online-G.0.en-fi.sgm"]=0.767
["newstest2017.TALP-UPC.4939.en-fi.sgm"]=0.806
["newstest2017.C-3MA.5069.en-lv.sgm"]=0.843
["newstest2017.HY-HNMT.5066.en-lv.sgm"]=0.786
["newstest2017.jhu-pbmt.4969.en-lv.sgm"]=0.815
["newstest2017.KIT.5062.en-lv.sgm"]=0.773
["newstest2017.limsi-factored-norm.5041.en-lv.sgm"]=0.780
["newstest2017.LIUM-FNMT.5043.en-lv.sgm"]=0.780
["newstest2017.LIUM-NMT.5042.en-lv.sgm"]=0.769
["newstest2017.online-A.0.en-lv.sgm"]=0.846
["newstest2017.online-B.0.en-lv.sgm"]=0.754
["newstest2017.PJATK.4744.en-lv.sgm"]=0.868
["newstest2017.tilde-c-nmt-smt-hybrid.5049.en-lv.sgm"]=0.731
["newstest2017.tilde-nc-nmt-smt-hybrid.5047.en-lv.sgm"]=0.721
["newstest2017.tilde-nc-smt.5044.en-lv.sgm"]=0.747
["newstest2017.uedin-nmt.5016.en-lv.sgm"]=0.772
["newstest2017.usfd-consensus-kit.5078.en-lv.sgm"]=0.950
["newstest2017.usfd-consensus-qt21.5077.en-lv.sgm"]=0.946
["newstest2017.afrl-mitll-backtrans.4907.en-ru.sgm"]=0.655
["newstest2017.jhu-pbmt.4986.en-ru.sgm"]=0.658
["newstest2017.online-A.0.en-ru.sgm"]=0.661
["newstest2017.online-B.0.en-ru.sgm"]=0.563
["newstest2017.online-F.0.en-ru.sgm"]=0.809
["newstest2017.online-G.0.en-ru.sgm"]=0.618
["newstest2017.PROMT-Rule-based.4736.en-ru.sgm"]=0.667
["newstest2017.uedin-nmt.4756.en-ru.sgm"]=0.614
["newstest2017.JAIST.4858.en-tr.sgm"]=0.875
["newstest2017.jhu-nmt-lattice-rescore.4904.en-tr.sgm"]=0.864
["newstest2017.jhu-pbmt.4970.en-tr.sgm"]=0.874
["newstest2017.LIUM-NMT.4953.en-tr.sgm"]=0.752
["newstest2017.online-A.0.en-tr.sgm"]=0.787
["newstest2017.online-B.0.en-tr.sgm"]=0.677
["newstest2017.uedin-nmt.4932.en-tr.sgm"]=0.747
["newstest2017.CASICT-DCU-NMT.5157.en-zh.sgm"]=0.999
["newstest2017.online-A.0.en-zh.sgm"]=2.936
["newstest2017.online-B.0.en-zh.sgm"]=1.092
["newstest2017.online-F.0.en-zh.sgm"]=1.114
["newstest2017.online-G.0.en-zh.sgm"]=1.158
["newstest2017.Oregon-State-University-S.5174.en-zh.sgm"]=5.622
["newstest2017.SogouKnowing-nmt.5131.en-zh.sgm"]=1.402
["newstest2017.uedin-nmt.5111.en-zh.sgm"]=1.334
["newstest2017.UU-HNMT.5134.en-zh.sgm"]=1.000
["newstest2017.xmunmt.5165.en-zh.sgm"]=1.366
)
if [ -z $SKIP_TER ]; then
echo "-------------------"
echo "Starting TER tests"
echo "-------------------"
for pair in cs-en en-cs en-de en-fi en-lv en-ru en-tr en-zh; do
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for sgm in wmt17-submitted-data/sgm/system-outputs/newstest2017/$pair/*.sgm; do
name=$(basename $sgm)
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
if [[ ! -v "TER[$name]" ]]; then continue ; fi # official Tercom fails for some outputs...
sys=$(basename $sgm .sgm | perl -pe 's/newstest2017\.//')
txt=$(dirname $sgm | perl -pe 's/sgm/txt/')/$(basename $sgm .sgm)
src=wmt17-submitted-data/sgm/sources/newstest2017-$source$target-src.$source.sgm
ref=wmt17-submitted-data/sgm/references/newstest2017-$source$target-ref.$target.sgm
score=$(cat $txt | ${CMD} -w 3 -t wmt17 -l $source-$target -b --metrics ter)
expected_score="${TER[$name]}"
echo "import sys; sys.exit(1 if abs(0.01 * ${score}-${expected_score}) > 0.01 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED test $pair/$sys (wanted $expected_score got $score)"
exit 1
fi
echo "Passed $source-$target $sys tercom: $expected_score sacreTER: $score"
let i++
done
done
fi
#############
# Mecab tests
#############
if [[ -z $SKIP_MECAB ]]; then
echo "-----------------------"
echo "Testing Mecab tokenizer"
echo "-----------------------"
pair="en-ja"
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for txt in en-ja-translation-example-master/*.hyp.$target; do
name=$(basename $txt .hyp.$target)
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
sys=$(basename $txt .hyp.$target)
ref=$(dirname $txt)/$(basename $txt .hyp.$target).ref.$target
score=$(cat $txt | ${CMD} -w 2 -l $source-$target -b $ref)
echo "import sys; sys.exit(1 if abs($score-${MTEVAL[$name]}) > 0.01 else 0)" | $PYTHON
if [[ $? -eq 1 ]]; then
echo "FAILED test $pair/$sys (wanted ${MTEVAL[$name]} got $score)"
exit 1
fi
echo "Passed $source-$target $sys mteval-v13a.pl: ${MTEVAL[$name]} sacreBLEU: $score"
let i++
done
fi
#############
# Mecab-ko tests
#############
if [[ -z $SKIP_MECAB_KO ]]; then
echo "-----------------------"
echo "Testing Mecab-ko tokenizer"
echo "-----------------------"
pair="en-ko"
source=$(echo $pair | cut -d- -f1)
target=$(echo $pair | cut -d- -f2)
for txt in en-ko-translation-example-master/*.hyp.$target; do
name=$(basename $txt .hyp.$target)
if [[ ! -z $limit_test && $limit_test != $name ]]; then continue; fi
sys=$(basename $txt .hyp.$target)
ref=$(dirname $txt)/$(basename $txt .hyp.$target).ref.$target
score=$(cat $txt | ${CMD} -w 2 -l $source-$target -b $ref)
echo "import sys; sys.exit(1 if abs($score-${MTEVAL[$name]}) > 0.01 else 0)" | python
if [[ $? -eq 1 ]]; then
echo "FAILED test $pair/$sys (wanted ${MTEVAL[$name]} got $score)"
exit 1
fi
echo "Passed $source-$target $sys mteval-v13a.pl: ${MTEVAL[$name]} sacreBLEU: $score"
let i++
done
fi
echo "Passed $i tests."
exit 0