Update README.md
Browse files
README.md
ADDED
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language: sla
|
3 |
+
tags:
|
4 |
+
- translation
|
5 |
+
|
6 |
+
license: apache-2.0
|
7 |
+
---
|
8 |
+
|
9 |
+
### sla-sla
|
10 |
+
|
11 |
+
* source languages: sla
|
12 |
+
* target languages: sla
|
13 |
+
* OPUS readme: [sla-sla](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/sla-sla/README.md)
|
14 |
+
|
15 |
+
* dataset: opus
|
16 |
+
* model: transformer
|
17 |
+
* source language(s): bel bel_Latn bos_Latn bul bul_Latn ces dsb hrv hsb mkd orv_Cyrl pol rus slv srp_Cyrl srp_Latn ukr
|
18 |
+
* target language(s): bel bel_Latn bos_Latn bul bul_Latn ces dsb hrv hsb mkd orv_Cyrl pol rus slv srp_Cyrl srp_Latn ukr
|
19 |
+
* model: transformer
|
20 |
+
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
|
21 |
+
* a sentence initial language token is required in the form of `>>id<<` (id = valid target language ID)
|
22 |
+
* download original weights: [opus-2020-07-27.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/sla-sla/opus-2020-07-27.zip)
|
23 |
+
* test set translations: [opus-2020-07-27.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/sla-sla/opus-2020-07-27.test.txt)
|
24 |
+
* test set scores: [opus-2020-07-27.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/sla-sla/opus-2020-07-27.eval.txt)
|
25 |
+
|
26 |
+
## Benchmarks
|
27 |
+
|
28 |
+
| testset | BLEU | chr-F |
|
29 |
+
|-----------------------|-------|-------|
|
30 |
+
| newstest2012-cesrus.ces.rus | 15.9 | 0.437 |
|
31 |
+
| newstest2012-rusces.rus.ces | 13.6 | 0.403 |
|
32 |
+
| newstest2013-cesrus.ces.rus | 19.8 | 0.473 |
|
33 |
+
| newstest2013-rusces.rus.ces | 17.9 | 0.449 |
|
34 |
+
| Tatoeba-test.bel-bul.bel.bul | 100.0 | 1.000 |
|
35 |
+
| Tatoeba-test.bel-ces.bel.ces | 33.5 | 0.630 |
|
36 |
+
| Tatoeba-test.bel-hbs.bel.hbs | 45.4 | 0.644 |
|
37 |
+
| Tatoeba-test.bel-mkd.bel.mkd | 19.3 | 0.531 |
|
38 |
+
| Tatoeba-test.bel-pol.bel.pol | 46.9 | 0.681 |
|
39 |
+
| Tatoeba-test.bel-rus.bel.rus | 58.5 | 0.767 |
|
40 |
+
| Tatoeba-test.bel-ukr.bel.ukr | 55.1 | 0.743 |
|
41 |
+
| Tatoeba-test.bul-bel.bul.bel | 10.7 | 0.423 |
|
42 |
+
| Tatoeba-test.bul-ces.bul.ces | 36.9 | 0.585 |
|
43 |
+
| Tatoeba-test.bul-hbs.bul.hbs | 53.7 | 0.807 |
|
44 |
+
| Tatoeba-test.bul-mkd.bul.mkd | 31.9 | 0.715 |
|
45 |
+
| Tatoeba-test.bul-pol.bul.pol | 38.6 | 0.607 |
|
46 |
+
| Tatoeba-test.bul-rus.bul.rus | 44.8 | 0.655 |
|
47 |
+
| Tatoeba-test.bul-ukr.bul.ukr | 49.9 | 0.691 |
|
48 |
+
| Tatoeba-test.ces-bel.ces.bel | 30.9 | 0.585 |
|
49 |
+
| Tatoeba-test.ces-bul.ces.bul | 75.8 | 0.859 |
|
50 |
+
| Tatoeba-test.ces-hbs.ces.hbs | 50.0 | 0.661 |
|
51 |
+
| Tatoeba-test.ces-hsb.ces.hsb | 7.9 | 0.246 |
|
52 |
+
| Tatoeba-test.ces-mkd.ces.mkd | 24.6 | 0.569 |
|
53 |
+
| Tatoeba-test.ces-pol.ces.pol | 44.3 | 0.652 |
|
54 |
+
| Tatoeba-test.ces-rus.ces.rus | 50.8 | 0.690 |
|
55 |
+
| Tatoeba-test.ces-slv.ces.slv | 4.9 | 0.240 |
|
56 |
+
| Tatoeba-test.ces-ukr.ces.ukr | 52.9 | 0.687 |
|
57 |
+
| Tatoeba-test.dsb-pol.dsb.pol | 16.3 | 0.367 |
|
58 |
+
| Tatoeba-test.dsb-rus.dsb.rus | 12.7 | 0.245 |
|
59 |
+
| Tatoeba-test.hbs-bel.hbs.bel | 32.9 | 0.531 |
|
60 |
+
| Tatoeba-test.hbs-bul.hbs.bul | 100.0 | 1.000 |
|
61 |
+
| Tatoeba-test.hbs-ces.hbs.ces | 40.3 | 0.626 |
|
62 |
+
| Tatoeba-test.hbs-mkd.hbs.mkd | 19.3 | 0.535 |
|
63 |
+
| Tatoeba-test.hbs-pol.hbs.pol | 45.0 | 0.650 |
|
64 |
+
| Tatoeba-test.hbs-rus.hbs.rus | 53.5 | 0.709 |
|
65 |
+
| Tatoeba-test.hbs-ukr.hbs.ukr | 50.7 | 0.684 |
|
66 |
+
| Tatoeba-test.hsb-ces.hsb.ces | 17.9 | 0.366 |
|
67 |
+
| Tatoeba-test.mkd-bel.mkd.bel | 23.6 | 0.548 |
|
68 |
+
| Tatoeba-test.mkd-bul.mkd.bul | 54.2 | 0.833 |
|
69 |
+
| Tatoeba-test.mkd-ces.mkd.ces | 12.1 | 0.371 |
|
70 |
+
| Tatoeba-test.mkd-hbs.mkd.hbs | 19.3 | 0.577 |
|
71 |
+
| Tatoeba-test.mkd-pol.mkd.pol | 53.7 | 0.833 |
|
72 |
+
| Tatoeba-test.mkd-rus.mkd.rus | 34.2 | 0.745 |
|
73 |
+
| Tatoeba-test.mkd-ukr.mkd.ukr | 42.7 | 0.708 |
|
74 |
+
| Tatoeba-test.multi.multi | 48.5 | 0.672 |
|
75 |
+
| Tatoeba-test.orv-pol.orv.pol | 10.1 | 0.355 |
|
76 |
+
| Tatoeba-test.orv-rus.orv.rus | 10.6 | 0.275 |
|
77 |
+
| Tatoeba-test.orv-ukr.orv.ukr | 7.5 | 0.230 |
|
78 |
+
| Tatoeba-test.pol-bel.pol.bel | 29.8 | 0.533 |
|
79 |
+
| Tatoeba-test.pol-bul.pol.bul | 36.8 | 0.578 |
|
80 |
+
| Tatoeba-test.pol-ces.pol.ces | 43.6 | 0.626 |
|
81 |
+
| Tatoeba-test.pol-dsb.pol.dsb | 0.9 | 0.097 |
|
82 |
+
| Tatoeba-test.pol-hbs.pol.hbs | 42.4 | 0.644 |
|
83 |
+
| Tatoeba-test.pol-mkd.pol.mkd | 19.3 | 0.535 |
|
84 |
+
| Tatoeba-test.pol-orv.pol.orv | 0.7 | 0.109 |
|
85 |
+
| Tatoeba-test.pol-rus.pol.rus | 49.6 | 0.680 |
|
86 |
+
| Tatoeba-test.pol-slv.pol.slv | 7.3 | 0.262 |
|
87 |
+
| Tatoeba-test.pol-ukr.pol.ukr | 46.8 | 0.664 |
|
88 |
+
| Tatoeba-test.rus-bel.rus.bel | 34.4 | 0.577 |
|
89 |
+
| Tatoeba-test.rus-bul.rus.bul | 45.5 | 0.657 |
|
90 |
+
| Tatoeba-test.rus-ces.rus.ces | 48.0 | 0.659 |
|
91 |
+
| Tatoeba-test.rus-dsb.rus.dsb | 10.7 | 0.029 |
|
92 |
+
| Tatoeba-test.rus-hbs.rus.hbs | 44.6 | 0.655 |
|
93 |
+
| Tatoeba-test.rus-mkd.rus.mkd | 34.9 | 0.617 |
|
94 |
+
| Tatoeba-test.rus-orv.rus.orv | 0.1 | 0.073 |
|
95 |
+
| Tatoeba-test.rus-pol.rus.pol | 45.2 | 0.659 |
|
96 |
+
| Tatoeba-test.rus-slv.rus.slv | 30.4 | 0.476 |
|
97 |
+
| Tatoeba-test.rus-ukr.rus.ukr | 57.6 | 0.751 |
|
98 |
+
| Tatoeba-test.slv-ces.slv.ces | 42.5 | 0.604 |
|
99 |
+
| Tatoeba-test.slv-pol.slv.pol | 39.6 | 0.601 |
|
100 |
+
| Tatoeba-test.slv-rus.slv.rus | 47.2 | 0.638 |
|
101 |
+
| Tatoeba-test.slv-ukr.slv.ukr | 36.4 | 0.549 |
|
102 |
+
| Tatoeba-test.ukr-bel.ukr.bel | 36.9 | 0.597 |
|
103 |
+
| Tatoeba-test.ukr-bul.ukr.bul | 56.4 | 0.733 |
|
104 |
+
| Tatoeba-test.ukr-ces.ukr.ces | 52.1 | 0.686 |
|
105 |
+
| Tatoeba-test.ukr-hbs.ukr.hbs | 47.1 | 0.670 |
|
106 |
+
| Tatoeba-test.ukr-mkd.ukr.mkd | 20.8 | 0.548 |
|
107 |
+
| Tatoeba-test.ukr-orv.ukr.orv | 0.2 | 0.058 |
|
108 |
+
| Tatoeba-test.ukr-pol.ukr.pol | 50.1 | 0.695 |
|
109 |
+
| Tatoeba-test.ukr-rus.ukr.rus | 63.9 | 0.790 |
|
110 |
+
| Tatoeba-test.ukr-slv.ukr.slv | 14.5 | 0.288 |
|
111 |
+
|