File size: 1,229 Bytes
deaef8a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: mit
---
**Model Description** 

Model created with OpenNMT-py 3.2 for the Spanish-Aragonese pair using a transformer architecture. The model was converted to the ctranslate2 format.
This model was trained for the paper Training and fine-tuning NMT models for low-resource languages using Apertium-based synthetic corpora

**How to Translate with this Model**

+ Install [Python 3.9](https://www.python.org/downloads/release/python-390/) 
+ Install [ctranslate 3.2](https://github.com/OpenNMT/CTranslate2)
+ Translate an input_text using the NOS-MT-es-arn model with the following command:
```bash
    perl tokenizer.perl < input.txt > input.tok
```
```bash
    subword_nmt.apply_bpe -c ./bpe/es.bpe < input.tok > input.bpe
```
```bash 
    python3 translate.py ./ct2-arn input.bpe > output.txt
```
```bash
    sed -i 's/@@ //g' output.txt
```

## Citation

If you use this model in your research, please cite the following paper:
Sant, A., Bardanca Outeiri帽o, D., Pichel Campos, J. R., De Luca Fornaciari, F., Escolano, C., Garc铆a Gilabert, J., Gamallo Otero, P., Mash, A., Liao, X., & Melero, M. (2023). Training and fine-tuning NMT models for low-resource languages using Apertium-based synthetic corpora. arXiv.