Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language: fa
|
3 |
+
tags:
|
4 |
+
- Style transfer
|
5 |
+
- Formality style transfer
|
6 |
+
widget:
|
7 |
+
- text: "من دوستام میرم بازی."
|
8 |
+
- text: "من به خونه دوستم رفتم."
|
9 |
+
|
10 |
+
---
|
11 |
+
|
12 |
+
# Persian-t5-formality-transfer
|
13 |
+
|
14 |
+
This is a formality style transfer model for the Persian language to convert colloquial text into a formal one. It is based on [the monolingual T5 model for Persian.](https://huggingface.co/Ahmad/parsT5-base) and [Persian T5 paraphraser](https://huggingface.co/erfan226/persian-t5-paraphraser)
|
15 |
+
|
16 |
+
## Usage
|
17 |
+
|
18 |
+
```python
|
19 |
+
|
20 |
+
>>> pip install transformers
|
21 |
+
>>> from transformers import (T5ForConditionalGeneration, AutoTokenizer, pipeline)
|
22 |
+
>>> import torch
|
23 |
+
|
24 |
+
model_path = 'erfan226/persian-t5-formality-transfer'
|
25 |
+
model = T5ForConditionalGeneration.from_pretrained(model_path)
|
26 |
+
tokenizer = AutoTokenizer.from_pretrained(model_path)
|
27 |
+
pipe = pipeline(task='text2text-generation', model=model, tokenizer=tokenizer)
|
28 |
+
|
29 |
+
def paraphrase(text):
|
30 |
+
for j in range(3):
|
31 |
+
out = pipe(text, encoder_no_repeat_ngram_size=4, do_sample=True, num_beams=5, max_length=128)[0]['generated_text']
|
32 |
+
print("Paraphrase:", out)
|
33 |
+
|
34 |
+
text = "من با دوستام میرم بازی"
|
35 |
+
print("Original:", text)
|
36 |
+
paraphrase(text)
|
37 |
+
|
38 |
+
# Original: من با دوستام میرم بازی
|
39 |
+
# Paraphrase: دوست دارم با دوستانم بازی کنم.
|
40 |
+
# Paraphrase: من با دوستانم میرم...
|
41 |
+
# Paraphrase: من با دوستام بازی می کنم.
|
42 |
+
|
43 |
+
```
|
44 |
+
|
45 |
+
## Training data
|
46 |
+
TBD
|