Translate Eng Cat

🚀 OpenNMT English - Catalan Translation Model

This is an English - Catalan translation model based on OpenNMT, which is also used in production at https://www.softcatala.org/traductor/.

🚀 Quick Start

This is an English - Catalan translation model based on OpenNMT. It offers reliable translation capabilities and is utilized in production at https://www.softcatala.org/traductor/.

📦 Installation

You can install the necessary dependencies using the following command:

pip3 install ctranslate2 pyonmttok

💻 Usage Examples

Basic Usage

Here is a simple Python example for translation:

import ctranslate2
import pyonmttok
from huggingface_hub import snapshot_download
model_dir = snapshot_download(repo_id="softcatala/translate-eng-cat", revision="main")

tokenizer=pyonmttok.Tokenizer(mode="none", sp_model_path = model_dir + "/sp_m.model")
tokenized=tokenizer.tokenize("Hello world!")

import ctranslate2
translator = ctranslate2.Translator(model_dir)
translated = translator.translate_batch([tokenized[0]])
print(tokenizer.detokenize(translated[0][0]['tokens']))
Hola món!

📚 Documentation

Benchmarks

The following table shows the BLEU scores on different test sets:

Additional information

• Softcatala NMT models
• Softcatala parallel Catalan corpus

📄 License

This project is licensed under the MIT license.