README.md

---
language: 
  - vi
thumbnail: "https://raw.githubusercontent.com/kldarek/polbert/master/img/polbert.png"
tags:
- transfomer
- sbert
- legaltext
- vietnamese
license: "mit"
datasets:
- vietnamese-legal-text
---

# Vietnamese Legal Text BERT
#### Table of contents
1. [Introduction](#introduction)
2. [Using Vietnamese Legal Text BERT](#transformers)
	- [Installation](#install2)
	- [Pre-trained models](#models2)
	- [Example usage](#usage2)

# <a name="introduction"></a> Using Vietnamese Legal Text BERT `hmthanh/VietnamLegalText-SBERT`


## <a name="transformers"></a> Using Vietnamese Legal Text BERT `transformers`

### Installation <a name="install2"></a>

- Install `transformers` with pip:

```pip install transformers```<br />

- Install `tokenizers` with pip:

```pip install tokenizers```

### Pre-trained models <a name="models2"></a>


Model | #params | Arch.	 | Max length | Pre-training data
---|---|---|---|---
`hmthanh/VietnamLegalText-SBERT` | 135M | base | 256 | 20GB  of texts

### Example usage <a name="usage2"></a>

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

phobert = AutoModel.from_pretrained("hmthanh/VietnamLegalText-SBERT")
tokenizer = AutoTokenizer.from_pretrained("hmthanh/VietnamLegalText-SBERT")

sentence = 'Vượt đèn đỏ bị phạt bao nhiêu tiền?'  

input_ids = torch.tensor([tokenizer.encode(sentence)])

with torch.no_grad():
    features = phobert(input_ids)  # Models outputs are now tuples
```