feat add readme
Browse files
README.md
ADDED
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- ru
|
4 |
+
- zh
|
5 |
+
- en
|
6 |
+
tags:
|
7 |
+
- translation
|
8 |
+
license: apache-2.0
|
9 |
+
datasets:
|
10 |
+
- ccmatrix
|
11 |
+
metrics:
|
12 |
+
- sacrebleu
|
13 |
+
---
|
14 |
+
|
15 |
+
# T5 English, Russian and Chinese multilingual machine translation
|
16 |
+
|
17 |
+
This model represents a conventional T5 transformer in multitasking mode for translation into the required language, precisely configured for machine translation for pairs: ru-zh, zh-ru, en-zh, zh-en, en-ru, ru-en.
|
18 |
+
|
19 |
+
The model can perform direct translation between any pair of Russian, Chinese or English languages. For translation into the target language, the target language identifier is specified as a prefix 'translate to <lang>:'. In this case, the source language may not be specified, in addition, the source text may be multilingual.
|
20 |
+
|
21 |
+
Example translate Russian to Chinese
|
22 |
+
|
23 |
+
```python
|
24 |
+
from transformers import T5ForConditionalGeneration, T5Tokenizer
|
25 |
+
|
26 |
+
model_name = 'utrobinmv/t5_translate_en_ru_zh_large_1024'
|
27 |
+
model = T5ForConditionalGeneration.from_pretrained(model_name)
|
28 |
+
tokenizer = T5Tokenizer.from_pretrained(model_name)
|
29 |
+
|
30 |
+
prefix = 'translate to zh: '
|
31 |
+
src_text = prefix + "Съешь ещё этих мягких французских булок."
|
32 |
+
|
33 |
+
# translate Russian to Chinese
|
34 |
+
input_ids = tokenizer(src_text, return_tensors="pt")
|
35 |
+
|
36 |
+
generated_tokens = model.generate(**input_ids)
|
37 |
+
|
38 |
+
result = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
|
39 |
+
print(result)
|
40 |
+
# 再吃这些法国的甜蜜的面包。
|
41 |
+
```
|
42 |
+
|
43 |
+
|
44 |
+
|
45 |
+
and Example translate Chinese to Russian
|
46 |
+
|
47 |
+
```python
|
48 |
+
from transformers import T5ForConditionalGeneration, T5Tokenizer
|
49 |
+
|
50 |
+
model_name = 'utrobinmv/t5_translate_en_ru_zh_large_1024'
|
51 |
+
model = T5ForConditionalGeneration.from_pretrained(model_name)
|
52 |
+
tokenizer = T5Tokenizer.from_pretrained(model_name)
|
53 |
+
|
54 |
+
prefix = 'translate to ru: '
|
55 |
+
src_text = prefix + "再吃这些法国的甜蜜的面包。"
|
56 |
+
|
57 |
+
# translate Russian to Chinese
|
58 |
+
input_ids = tokenizer(src_text, return_tensors="pt")
|
59 |
+
|
60 |
+
generated_tokens = model.generate(**input_ids)
|
61 |
+
|
62 |
+
result = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
|
63 |
+
print(result)
|
64 |
+
# Съешьте этот сладкий хлеб из Франции.
|
65 |
+
```
|
66 |
+
|
67 |
+
|
68 |
+
|
69 |
+
##
|
70 |
+
|
71 |
+
|
72 |
+
|
73 |
+
## Languages covered
|
74 |
+
|
75 |
+
Russian (ru_RU), Chinese (zh_CN), English (en_US)
|