Fugumt En Ja

🚀 FuguMT

FuguMT is a translation model built on Marian-NMT. It enables English-to-Japanese translation. For more in - depth information, visit my repository.

🚀 Quick Start

✨ Features

• Source language: English
• Target language: Japanese

📦 Installation

This model relies on transformers and sentencepiece. Install them using the following command:

!pip install transformers sentencepiece

💻 Usage Examples

Basic Usage

You can directly use this model with a pipeline:

from transformers import pipeline
fugu_translator = pipeline('translation', model='staka/fugumt-en-ja')
fugu_translator('This is a cat.')

Advanced Usage

If you need to translate multiple sentences, we recommend using pySBD.

!pip install transformers sentencepiece pysbd

import pysbd
seg_en = pysbd.Segmenter(language="en", clean=False)

from transformers import pipeline
fugu_translator = pipeline('translation', model='staka/fugumt-en-ja')
txt = 'This is a cat. It is very cute.'
print(fugu_translator(seg_en.segment(txt)))

📚 Documentation

Eval results

The evaluation results using tatoeba (randomly selected 500 sentences) are presented below:

(*1) sacrebleu --tokenize ja-mecab

📄 License

This project is licensed under CC - BY - SA 4.0.