From b4a905969a633dd5885198b8949e71133627473e Mon Sep 17 00:00:00 2001 From: neurlang <77860779+neurlang@users.noreply.github.com> Date: Sat, 8 Feb 2025 22:06:27 +0100 Subject: [PATCH] add experimental results --- coverage_forward.txt | 85 ++++++++++++++++++++++++++++++++++++++++++++ coverage_reverse.txt | 85 ++++++++++++++++++++++++++++++++++++++++++++ success_forward.txt | 85 ++++++++++++++++++++++++++++++++++++++++++++ success_reverse.txt | 85 ++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 340 insertions(+) create mode 100644 coverage_forward.txt create mode 100644 coverage_reverse.txt create mode 100644 success_forward.txt create mode 100644 success_reverse.txt diff --git a/coverage_forward.txt b/coverage_forward.txt new file mode 100644 index 0000000..e1dc519 --- /dev/null +++ b/coverage_forward.txt @@ -0,0 +1,85 @@ +Coverage forward: 137% for language chinese/mandarin +Coverage forward: 18% for language hebrew +Coverage forward: 19% for language tibetan +Coverage forward: 22% for language norwegian +Coverage forward: 22% for language telugu +Coverage forward: 23% for language english +Coverage forward: 24% for language afrikaans +Coverage forward: 24% for language arabic +Coverage forward: 28% for language greek +Coverage forward: 28% for language tamil +Coverage forward: 30% for language malayalam +Coverage forward: 31% for language danish +Coverage forward: 31% for language finnish +Coverage forward: 32% for language italian +Coverage forward: 32% for language urdu +Coverage forward: 34% for language german +Coverage forward: 36% for language javanese +Coverage forward: 36% for language pashto +Coverage forward: 37% for language spanish +Coverage forward: 38% for language dutch +Coverage forward: 38% for language punjabi +Coverage forward: 39% for language portuguese +Coverage forward: 39% for language turkish +Coverage forward: 39% for language vietnamese/northern +Coverage forward: 40% for language malay/latin +Coverage forward: 41% for language gujarati +Coverage forward: 42% for language luxembourgish +Coverage forward: 44% for language french +Coverage forward: 45% for language bengali/dhaka +Coverage forward: 45% for language farsi +Coverage forward: 45% for language marathi +Coverage forward: 45% for language thai +Coverage forward: 47% for language bengali/rahr +Coverage forward: 48% for language indonesian +Coverage forward: 52% for language swedish +Coverage forward: 53% for language kazakh +Coverage forward: 54% for language bengali +Coverage forward: 54% for language vietnamese/central +Coverage forward: 54% for language vietnamese/southern +Coverage forward: 56% for language burmese +Coverage forward: 56% for language romanian +Coverage forward: 57% for language ukrainian +Coverage forward: 62% for language hindi +Coverage forward: 62% for language icelandic +Coverage forward: 64% for language malay/arab +Coverage forward: 67% for language hungarian +Coverage forward: 68% for language cebuano +Coverage forward: 69% for language czech +Coverage forward: 69% for language tagalog +Coverage forward: 70% for language nepali +Coverage forward: 71% for language japanese +Coverage forward: 73% for language macedonian +Coverage forward: 74% for language armenian +Coverage forward: 75% for language lao +Coverage forward: 75% for language maltese +Coverage forward: 76% for language korean +Coverage forward: 77% for language azerbaijani +Coverage forward: 78% for language mongolian +Coverage forward: 79% for language basque +Coverage forward: 80% for language estonian +Coverage forward: 81% for language isan +Coverage forward: 83% for language khmer +Coverage forward: 84% for language hausa +Coverage forward: 86% for language amharic +Coverage forward: 86% for language chechen +Coverage forward: 86% for language russian +Coverage forward: 87% for language swahili +Coverage forward: 89% for language lithuanian +Coverage forward: 89% for language polish +Coverage forward: 90% for language belarusian +Coverage forward: 90% for language zulu +Coverage forward: 91% for language uyghur +Coverage forward: 92% for language galician +Coverage forward: 94% for language croatian +Coverage forward: 94% for language yoruba +Coverage forward: 95% for language catalan +Coverage forward: 95% for language dzongkha +Coverage forward: 95% for language latvian +Coverage forward: 97% for language chichewa +Coverage forward: 97% for language jamaican +Coverage forward: 97% for language serbian +Coverage forward: 97% for language slovak +Coverage forward: 99% for language bulgarian +Coverage forward: 99% for language esperanto +Coverage forward: 99% for language georgian diff --git a/coverage_reverse.txt b/coverage_reverse.txt new file mode 100644 index 0000000..6362185 --- /dev/null +++ b/coverage_reverse.txt @@ -0,0 +1,85 @@ +Coverage reverse: 1% for language tibetan +Coverage reverse: 16% for language arabic +Coverage reverse: 28% for language french +Coverage reverse: 30% for language hebrew +Coverage reverse: 32% for language danish +Coverage reverse: 33% for language malayalam +Coverage reverse: 41% for language english +Coverage reverse: 41% for language lithuanian +Coverage reverse: 41% for language pashto +Coverage reverse: 42% for language urdu +Coverage reverse: 43% for language german +Coverage reverse: 44% for language afrikaans +Coverage reverse: 44% for language dutch +Coverage reverse: 44% for language tamil +Coverage reverse: 44% for language vietnamese/northern +Coverage reverse: 47% for language norwegian +Coverage reverse: 52% for language kazakh +Coverage reverse: 52% for language luxembourgish +Coverage reverse: 52% for language marathi +Coverage reverse: 52% for language punjabi +Coverage reverse: 54% for language turkish +Coverage reverse: 56% for language malay/latin +Coverage reverse: 57% for language icelandic +Coverage reverse: 60% for language gujarati +Coverage reverse: 61% for language greek +Coverage reverse: 65% for language ukrainian +Coverage reverse: 66% for language farsi +Coverage reverse: 66% for language indonesian +Coverage reverse: 66% for language thai +Coverage reverse: 66% for language vietnamese/central +Coverage reverse: 67% for language cebuano +Coverage reverse: 67% for language javanese +Coverage reverse: 68% for language hungarian +Coverage reverse: 69% for language malay/arab +Coverage reverse: 69% for language mongolian +Coverage reverse: 69% for language romanian +Coverage reverse: 69% for language tagalog +Coverage reverse: 70% for language bengali/dhaka +Coverage reverse: 71% for language telugu +Coverage reverse: 73% for language swedish +Coverage reverse: 76% for language bengali/rahr +Coverage reverse: 76% for language burmese +Coverage reverse: 76% for language japanese +Coverage reverse: 77% for language bengali +Coverage reverse: 77% for language maltese +Coverage reverse: 78% for language isan +Coverage reverse: 78% for language yoruba +Coverage reverse: 79% for language azerbaijani +Coverage reverse: 80% for language nepali +Coverage reverse: 80% for language vietnamese/southern +Coverage reverse: 81% for language khmer +Coverage reverse: 81% for language macedonian +Coverage reverse: 82% for language portuguese +Coverage reverse: 84% for language basque +Coverage reverse: 84% for language hausa +Coverage reverse: 84% for language russian +Coverage reverse: 85% for language hindi +Coverage reverse: 86% for language chechen +Coverage reverse: 87% for language swahili +Coverage reverse: 88% for language armenian +Coverage reverse: 89% for language czech +Coverage reverse: 90% for language amharic +Coverage reverse: 90% for language galician +Coverage reverse: 90% for language korean +Coverage reverse: 90% for language polish +Coverage reverse: 92% for language italian +Coverage reverse: 92% for language spanish +Coverage reverse: 93% for language chinese/mandarin +Coverage reverse: 94% for language belarusian +Coverage reverse: 94% for language estonian +Coverage reverse: 94% for language uyghur +Coverage reverse: 95% for language catalan +Coverage reverse: 95% for language chichewa +Coverage reverse: 95% for language latvian +Coverage reverse: 95% for language zulu +Coverage reverse: 96% for language slovak +Coverage reverse: 97% for language dzongkha +Coverage reverse: 97% for language lao +Coverage reverse: 98% for language jamaican +Coverage reverse: 99% for language bulgarian +Coverage reverse: 99% for language croatian +Coverage reverse: 99% for language esperanto +Coverage reverse: 99% for language finnish +Coverage reverse: 99% for language georgian +Coverage reverse: 99% for language serbian diff --git a/success_forward.txt b/success_forward.txt new file mode 100644 index 0000000..b2442a6 --- /dev/null +++ b/success_forward.txt @@ -0,0 +1,85 @@ +[success rate] 00 % with 120245 errors 0 successes for japanese +[success rate] 00 % with 1217 errors 0 successes for malay/arab +[success rate] 00 % with 130866 errors 33 successes for tibetan +[success rate] 00 % with 38549 errors 3 successes for luxembourgish +[success rate] 00 % with 71971 errors 0 successes for malay/latin +[success rate] 00 % with 8918 errors 0 successes for hungarian +[success rate] 00 % with 99298 errors 0 successes for malayalam +[success rate] 01 % with 11537 errors 11 successes for spanish +[success rate] 01 % with 15749 errors 177 successes for serbian +[success rate] 01 % with 33353 errors 191 successes for hebrew +[success rate] 01 % with 6084 errors 27 successes for zulu +[success rate] 01 % with 69634 errors 116 successes for tamil +[success rate] 02 % with 45341 errors 259 successes for marathi +[success rate] 03 % with 112099 errors 341 successes for thai +[success rate] 03 % with 16098 errors 339 successes for croatian +[success rate] 03 % with 25647 errors 356 successes for russian +[success rate] 03 % with 37495 errors 397 successes for gujarati +[success rate] 03 % with 43006 errors 372 successes for german +[success rate] 04 % with 267859 errors 447 successes for chinese/mandarin +[success rate] 04 % with 38506 errors 425 successes for english +[success rate] 04 % with 51947 errors 435 successes for arabic +[success rate] 04 % with 6398 errors 78 successes for hausa +[success rate] 05 % with 31758 errors 544 successes for norwegian +[success rate] 05 % with 31823 errors 528 successes for afrikaans +[success rate] 06 % with 28940 errors 510 successes for danish +[success rate] 06 % with 29140 errors 609 successes for portuguese +[success rate] 06 % with 29367 errors 674 successes for urdu +[success rate] 06 % with 41327 errors 694 successes for telugu +[success rate] 07 % with 28241 errors 736 successes for ukrainian +[success rate] 07 % with 31443 errors 775 successes for punjabi +[success rate] 09 % with 22098 errors 930 successes for greek +[success rate] 09 % with 25194 errors 962 successes for turkish +[success rate] 09 % with 32951 errors 996 successes for dutch +[success rate] 10 % with 12038 errors 557 successes for yoruba +[success rate] 11 % with 17988 errors 723 successes for bengali/rahr +[success rate] 11 % with 18146 errors 1180 successes for italian +[success rate] 11 % with 20112 errors 893 successes for bengali/dhaka +[success rate] 11 % with 23831 errors 1108 successes for indonesian +[success rate] 11 % with 23839 errors 1181 successes for french +[success rate] 11 % with 24901 errors 1139 successes for icelandic +[success rate] 12 % with 14426 errors 749 successes for bengali +[success rate] 13 % with 2361 errors 112 successes for mongolian +[success rate] 13 % with 4251 errors 393 successes for cebuano +[success rate] 14 % with 24170 errors 1420 successes for hindi +[success rate] 14 % with 64435 errors 1446 successes for finnish +[success rate] 15 % with 13674 errors 937 successes for kazakh +[success rate] 15 % with 27263 errors 1598 successes for swedish +[success rate] 17 % with 15554 errors 1704 successes for tagalog +[success rate] 18 % with 1086 errors 125 successes for javanese +[success rate] 18 % with 53349 errors 1801 successes for farsi +[success rate] 19 % with 17177 errors 1207 successes for burmese +[success rate] 20 % with 2629 errors 267 successes for pashto +[success rate] 22 % with 4776 errors 226 successes for khmer/central +[success rate] 23 % with 13409 errors 1489 successes for vietnamese/northern +[success rate] 23 % with 53042 errors 2319 successes for korean +[success rate] 25 % with 25249 errors 1053 successes for lao +[success rate] 31 % with 39770 errors 3165 successes for macedonian +[success rate] 36 % with 12392 errors 3619 successes for maltese +[success rate] 37 % with 266 errors 65 successes for chechen +[success rate] 37 % with 3000 errors 723 successes for nepali +[success rate] 39 % with 11562 errors 2438 successes for vietnamese/central +[success rate] 40 % with 9528 errors 2356 successes for vietnamese/southern +[success rate] 42 % with 22973 errors 3379 successes for basque +[success rate] 42 % with 9414 errors 4291 successes for bulgarian +[success rate] 44 % with 10099 errors 1869 successes for galician +[success rate] 44 % with 11478 errors 4422 successes for romanian +[success rate] 46 % with 12740 errors 4630 successes for swahili +[success rate] 46 % with 7333 errors 4688 successes for czech +[success rate] 48 % with 11640 errors 4894 successes for catalan +[success rate] 48 % with 3459 errors 1943 successes for azerbaijani +[success rate] 50 % with 2213 errors 1112 successes for isan +[success rate] 50 % with 8771 errors 5030 successes for armenian +[success rate] 51 % with 47325 errors 5152 successes for lithuanian +[success rate] 54 % with 1015 errors 166 successes for dzongkha +[success rate] 55 % with 1914 errors 628 successes for estonian +[success rate] 55 % with 566 errors 243 successes for amharic +[success rate] 56 % with 36307 errors 5695 successes for georgian +[success rate] 60 % with 1015 errors 760 successes for latvian +[success rate] 61 % with 4901 errors 6198 successes for slovak +[success rate] 66 % with 5598 errors 6605 successes for polish +[success rate] 68 % with 1014 errors 569 successes for chichewa +[success rate] 71 % with 2149 errors 3950 successes for belarusian +[success rate] 75 % with 571 errors 1317 successes for jamaican +[success rate] 86 % with 348 errors 1227 successes for uyghur +[success rate] 99 % with 259 errors 9960 successes for esperanto diff --git a/success_reverse.txt b/success_reverse.txt new file mode 100644 index 0000000..fec9cae --- /dev/null +++ b/success_reverse.txt @@ -0,0 +1,85 @@ +[success rate] 00 % with 101014 errors 0 successes for malayalam +[success rate] 00 % with 1057 errors 0 successes for malay/arab +[success rate] 00 % with 193644 errors 49 successes for tibetan +[success rate] 00 % with 72434 errors 0 successes for malay/latin +[success rate] 01 % with 30340 errors 139 successes for english +[success rate] 01 % with 36476 errors 149 successes for french +[success rate] 02 % with 33568 errors 298 successes for arabic +[success rate] 03 % with 13930 errors 240 successes for vietnamese/northern +[success rate] 03 % with 7507 errors 110 successes for cebuano +[success rate] 04 % with 25053 errors 483 successes for greek +[success rate] 07 % with 21645 errors 713 successes for hebrew +[success rate] 07 % with 39830 errors 749 successes for chinese/mandarin +[success rate] 11 % with 22654 errors 1158 successes for afrikaans +[success rate] 12 % with 35906 errors 1277 successes for tamil +[success rate] 13 % with 29042 errors 1392 successes for hungarian +[success rate] 14 % with 23022 errors 1465 successes for icelandic +[success rate] 15 % with 15073 errors 1332 successes for danish +[success rate] 16 % with 20741 errors 1676 successes for norwegian +[success rate] 18 % with 57293 errors 1861 successes for thai +[success rate] 19 % with 23710 errors 1904 successes for marathi +[success rate] 21 % with 18472 errors 2136 successes for punjabi +[success rate] 22 % with 15171 errors 2182 successes for urdu +[success rate] 22 % with 17674 errors 2213 successes for german +[success rate] 22 % with 18209 errors 2212 successes for dutch +[success rate] 25 % with 16493 errors 2518 successes for luxembourgish +[success rate] 25 % with 17377 errors 1540 successes for bengali +[success rate] 26 % with 26312 errors 2635 successes for japanese +[success rate] 27 % with 17788 errors 2710 successes for gujarati +[success rate] 27 % with 2338 errors 362 successes for pashto +[success rate] 28 % with 33309 errors 2881 successes for telugu +[success rate] 29 % with 10341 errors 1801 successes for burmese +[success rate] 32 % with 7397 errors 1975 successes for kazakh +[success rate] 33 % with 12453 errors 2657 successes for bengali/dhaka +[success rate] 33 % with 13569 errors 3378 successes for russian +[success rate] 33 % with 15701 errors 3318 successes for korean +[success rate] 33 % with 17302 errors 3357 successes for lithuanian +[success rate] 33 % with 2981 errors 733 successes for isan +[success rate] 34 % with 8944 errors 2171 successes for khmer/central +[success rate] 35 % with 17822 errors 3530 successes for farsi +[success rate] 38 % with 13711 errors 3847 successes for turkish +[success rate] 38 % with 21257 errors 3818 successes for hindi +[success rate] 39 % with 11612 errors 3910 successes for ukrainian +[success rate] 40 % with 6299 errors 2553 successes for vietnamese/central +[success rate] 42 % with 10681 errors 4290 successes for indonesian +[success rate] 43 % with 9005 errors 4355 successes for portuguese +[success rate] 44 % with 12312 errors 4401 successes for slovak +[success rate] 44 % with 1364 errors 295 successes for javanese +[success rate] 46 % with 10148 errors 4692 successes for spanish +[success rate] 46 % with 9310 errors 3010 successes for bengali/rahr +[success rate] 46 % with 9433 errors 4693 successes for polish +[success rate] 47 % with 4827 errors 2770 successes for vietnamese/southern +[success rate] 48 % with 7214 errors 4846 successes for tagalog +[success rate] 48 % with 9412 errors 4881 successes for maltese +[success rate] 50 % with 10426 errors 5088 successes for catalan +[success rate] 50 % with 704 errors 152 successes for dzongkha +[success rate] 51 % with 2260 errors 995 successes for nepali +[success rate] 52 % with 8848 errors 5233 successes for czech +[success rate] 53 % with 294 errors 92 successes for chechen +[success rate] 55 % with 6221 errors 5555 successes for swedish +[success rate] 57 % with 2794 errors 2316 successes for azerbaijani +[success rate] 57 % with 6948 errors 5771 successes for romanian +[success rate] 60 % with 647 errors 489 successes for mongolian +[success rate] 62 % with 4697 errors 6226 successes for bulgarian +[success rate] 64 % with 5176 errors 6440 successes for armenian +[success rate] 64 % with 8408 errors 2676 successes for lao +[success rate] 67 % with 1814 errors 2858 successes for galician +[success rate] 67 % with 1815 errors 841 successes for latvian +[success rate] 70 % with 2926 errors 5601 successes for basque +[success rate] 70 % with 7995 errors 7000 successes for macedonian +[success rate] 71 % with 5116 errors 7154 successes for swahili +[success rate] 72 % with 1941 errors 3699 successes for yoruba +[success rate] 75 % with 376 errors 628 successes for chichewa +[success rate] 75 % with 648 errors 1445 successes for hausa +[success rate] 82 % with 162 errors 361 successes for amharic +[success rate] 82 % with 3532 errors 8207 successes for georgian +[success rate] 82 % with 4406 errors 8246 successes for serbian +[success rate] 83 % with 423 errors 1181 successes for uyghur +[success rate] 84 % with 1936 errors 8454 successes for italian +[success rate] 84 % with 2505 errors 8419 successes for esperanto +[success rate] 84 % with 978 errors 4675 successes for belarusian +[success rate] 88 % with 286 errors 1536 successes for zulu +[success rate] 89 % with 154 errors 1005 successes for estonian +[success rate] 92 % with 7038 errors 9216 successes for finnish +[success rate] 95 % with 134 errors 1670 successes for jamaican +[success rate] 97 % with 1070 errors 9745 successes for croatian