opus-mt-inc-inc / README.md
system
Update README.md 9100ff0
1
---
2
language: 
3
- bn
4
- or
5
- gu
6
- mr
7
- ur
8
- hi
9
- as
10
- si
11
- inc
12
13
tags:
14
- translation
15
16
license: apache-2.0
17
---
18
19
### inc-inc
20
21
* source group: Indic languages 
22
* target group: Indic languages 
23
*  OPUS readme: [inc-inc](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/inc-inc/README.md)
24
25
*  model: transformer
26
* source language(s): asm hin mar urd
27
* target language(s): asm hin mar urd
28
* model: transformer
29
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
30
* a sentence initial language token is required in the form of `>>id<<` (id = valid target language ID)
31
* download original weights: [opus-2020-07-27.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/inc-inc/opus-2020-07-27.zip)
32
* test set translations: [opus-2020-07-27.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/inc-inc/opus-2020-07-27.test.txt)
33
* test set scores: [opus-2020-07-27.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/inc-inc/opus-2020-07-27.eval.txt)
34
35
## Benchmarks
36
37
| testset               | BLEU  | chr-F |
38
|-----------------------|-------|-------|
39
| Tatoeba-test.asm-hin.asm.hin 	| 2.6 	| 0.231 |
40
| Tatoeba-test.hin-asm.hin.asm 	| 9.1 	| 0.262 |
41
| Tatoeba-test.hin-mar.hin.mar 	| 28.1 	| 0.548 |
42
| Tatoeba-test.hin-urd.hin.urd 	| 19.9 	| 0.508 |
43
| Tatoeba-test.mar-hin.mar.hin 	| 11.6 	| 0.466 |
44
| Tatoeba-test.multi.multi 	| 17.1 	| 0.464 |
45
| Tatoeba-test.urd-hin.urd.hin 	| 13.5 	| 0.377 |
46
47
48
### System Info: 
49
- hf_name: inc-inc
50
51
- source_languages: inc
52
53
- target_languages: inc
54
55
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/inc-inc/README.md
56
57
- original_repo: Tatoeba-Challenge
58
59
- tags: ['translation']
60
61
- languages: ['bn', 'or', 'gu', 'mr', 'ur', 'hi', 'as', 'si', 'inc']
62
63
- src_constituents: {'pnb', 'gom', 'ben', 'hif_Latn', 'ori', 'guj', 'pan_Guru', 'snd_Arab', 'npi', 'mar', 'urd', 'bho', 'hin', 'san_Deva', 'asm', 'rom', 'mai', 'awa', 'sin'}
64
65
- tgt_constituents: {'pnb', 'gom', 'ben', 'hif_Latn', 'ori', 'guj', 'pan_Guru', 'snd_Arab', 'npi', 'mar', 'urd', 'bho', 'hin', 'san_Deva', 'asm', 'rom', 'mai', 'awa', 'sin'}
66
67
- src_multilingual: True
68
69
- tgt_multilingual: True
70
71
- prepro:  normalization + SentencePiece (spm32k,spm32k)
72
73
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/inc-inc/opus-2020-07-27.zip
74
75
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/inc-inc/opus-2020-07-27.test.txt
76
77
- src_alpha3: inc
78
79
- tgt_alpha3: inc
80
81
- short_pair: inc-inc
82
83
- chrF2_score: 0.46399999999999997
84
85
- bleu: 17.1
86
87
- brevity_penalty: 1.0
88
89
- ref_len: 4985.0
90
91
- src_name: Indic languages
92
93
- tgt_name: Indic languages
94
95
- train_date: 2020-07-27
96
97
- src_alpha2: inc
98
99
- tgt_alpha2: inc
100
101
- prefer_old: False
102
103
- long_pair: inc-inc
104
105
- helsinki_git_sha: 480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535
106
107
- transformers_git_sha: 2207e5d8cb224e954a7cba69fa4ac2309e9ff30b
108
109
- port_machine: brutasse
110
111
- port_time: 2020-08-21-14:41