---
license: cc-by-nc-sa-4.0
language:
- ru
- uk
- es
- en
tags:
- detoxification
---

**Model Overview**

This is the model presented in the paper ["MultiParaDetox: Extending Text Detoxification with Parallel Data to New Languages"](https://arxiv.org/pdf/2404.02037). 
It is [facebook/mbart-large-50](https://huggingface.co/facebook/mbart-large-50) fine-tuned on the parallel detoxification dataset of RUssian, English, Ukrainian, and Spanish.

**How to use**
```python
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
model_name = 'textdetox/mBART_paradetox_rus_ukr_esp_en'
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
```

**Citation**
```
@article{dementieva2024multiparadetox,
  title={MultiParaDetox: Extending Text Detoxification with Parallel Data to New Languages},
  author={Dementieva, Daryna and Babakov, Nikolay and Panchenko, Alexander},
  journal={arXiv preprint arXiv:2404.02037},
  year={2024}
}
```