File size: 1,146 Bytes
2fda366
1b55c8a
054affb
1b55c8a
c1d685f
779f6a7
c1d685f
779f6a7
c1d685f
779f6a7
 
 
55175ab
c1d685f
 
8cade7d
 
 
 
 
 
 
 
2d90b23
8cade7d
 
 
 
 
 
 
 
 
 
 
 
 
c1d685f
b1d6579
1b55c8a
0310dda
1b55c8a
b1d6579
1b55c8a
0310dda
1b55c8a
c1d685f
8cade7d
 
 
0310dda
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
---
language:
- fa
license: mit
---
***This model is intended for non-commercial use only. If you wish to use it commercially, please refer to this LinkedIn address.

https://www.linkedin.com/in/amin-rahmani-41417b121/

"Viravirast" is an editor based on transformer algorithms. By visiting Viravirast.com, you can use a Persian semantic and structural text editor.


***Use this code in order to get the complete and correct sentence


from transformers import (
  T5Tokenizer,
  MT5ForConditionalGeneration,
  Text2TextGenerationPipeline,
)


path = ""

pipe = Text2TextGenerationPipeline(
  model=MT5ForConditionalGeneration.from_pretrained(path),
  tokenizer=T5Tokenizer.from_pretrained(path),
)


sentence = "ویراویراست یک نرم افزار ویرایش متن ساختاری و معنایی زبان فارسی است چیزی شبیه به گرامرلی در زبان انگلیسی"

#res = pipe(sentence, max_length=100, num_beams=4

res = pipe(sentence, max_length=100)

print(res[0]['generated_text'])


n_epochs = 4

train_batch_size = 8

eval_batch_size = 4

lr = 5e-4 

Training_Loss=0.00550

Validation_Loss=0.052046