File size: 2,173 Bytes
7d29d04
 
 
 
 
 
 
714ef19
7d29d04
 
 
 
 
 
6b5feee
 
 
 
7d29d04
 
 
 
 
 
 
 
208a846
 
 
 
 
 
 
 
 
 
 
 
 
 
a6b4b15
 
208a846
 
 
 
 
 
 
a6b4b15
208a846
7d29d04
 
 
 
 
 
 
 
 
 
a6b4b15
 
cd1ccf1
7d29d04
 
 
a6b4b15
 
 
 
7d29d04
 
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
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
---
license: apache-2.0
datasets:
- unicamp-dl/mmarco
language:
- vi
library_name: sentence-transformers
pipeline_tag: text-classification
tags:
- cross-encoder
- rerank
---

## Installation
  - Install `pyvi` to word segment:

	- `pip install pyvi`
      
 -  Install `sentence-transformers` (recommend):
	
	- `pip install sentence-transformers`

 -  Install `transformers` (optional):

	- `pip install transformers`

## Pre-processing

```python
from pyvi import ViTokenizer

query = "UIT là gì?"
sentences = [
    "UIT là Trường Đại học Công nghệ Thông tin (ĐH CNTT), Đại học Quốc gia Thành phố Hồ Chí Minh (ĐHQG-HCM)",
    "Mô hình rerank — còn được gọi là cross-encoder — là một loại mô hình mà, khi được cung cấp một cặp truy vấn và tài liệu, sẽ đưa ra một điểm tương đồng.",
    "Việt Nam, quốc hiệu là Cộng hòa xã hội chủ nghĩa Việt Nam, là một quốc gia xã hội chủ nghĩa nằm ở cực Đông của bán đảo Đông Dương thuộc khu vực Đông Nam Á"
]

tokenized_query = ViTokenizer.tokenize(query)
tokenized_sentences = [ViTokenizer.tokenize(sent) for sent in sentences]

tokenized_pairs = [[tokenized_query, sent] for sent in tokenized_sentences]
```

## Usage with sentence-transformers

```python
from sentence_transformers import CrossEncoder
model = CrossEncoder('itdainb/vietnamese-cross-encoder', max_length=256)
scores = model.predict(tokenized_pairs)
```

## Usage with transformers

```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model = AutoModelForSequenceClassification.from_pretrained('itdainb/vietnamese-cross-encoder')
tokenizer = AutoTokenizer.from_pretrained('itdainb/vietnamese-cross-encoder')

activation_fct = torch.nn.Identity()

features = tokenizer(*tokenized_pairs, padding=True, truncation="longest_first", return_tensors="pt", max_length=256)

model.eval()
with torch.no_grad():
    model_predictions = self.model(**features, return_dict=True)
    logits = activation_fct(model_predictions.logits)

    scores = [score[0] for score in logits]
    print(scores)
```