Opus-mt-ja-bg Open-source Machine Translation Model - Free and Precise Translation from Japanese to Bulgarian

Opus Mt Ja Bg

Developed by Helsinki-NLP

This is a Japanese-to-Bulgarian machine translation model based on the transformer-align architecture, developed by the Helsinki-NLP team.

Machine Translation

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Japanese-Bulgarian translation #Multiscript support #Low-resource language pair

Downloads 14

Release Time : 3/2/2022

Model Overview

This model is specifically designed for translating Japanese text into Bulgarian, utilizing the transformer-align architecture and SentencePiece preprocessing technology.

Model Features

Multiscript Support

Supports input in various Japanese writing systems (hiragana, katakana, kanji, etc.)

Standardized Preprocessing

Employs normalization and SentencePiece (spm32k) preprocessing techniques to enhance translation quality

Alignment Mechanism

Uses transformer-align architecture to improve source-target language alignment

Model Capabilities

Japanese-to-Bulgarian text translation

Multiscript Japanese processing

Cross-language alignment

Use Cases

Language Services

Document Translation

Translate Japanese documents into Bulgarian

Achieved 20.2 BLEU score on Tatoeba test set

Cross-language Communication

Facilitate communication between Japanese and Bulgarian speakers

🚀 JPN-BUL Translation Model

This project focuses on the translation between Japanese and Bulgarian. It provides a transformer - based model with specific pre - processing steps and detailed benchmark scores.

✨ Features

Language Pair: Translates from Japanese to Bulgarian.
Model Type: Uses the transformer - align model.
Pre - processing: Applies normalization and SentencePiece (spm32k, spm32k).

📦 Installation

The original weights of the model can be downloaded from opus - 2020 - 06 - 17.zip.

📚 Documentation

Model Details

Source Group: Japanese
Target Group: Bulgarian
OPUS Readme: jpn - bul
Model: transformer - align
Source Languages: jpn, jpn_Hani, jpn_Hira, jpn_Kana
Target Language: bul
Pre - processing: normalization + SentencePiece (spm32k, spm32k)

Download Links

Original Weights: opus - 2020 - 06 - 17.zip
Test Set Translations: opus - 2020 - 06 - 17.test.txt
Test Set Scores: opus - 2020 - 06 - 17.eval.txt

Benchmarks

testset	BLEU	chr - F
Tatoeba - test.jpn.bul	20.2	0.422

System Info

Property	Details
hf_name	jpn - bul
source_languages	jpn
target_languages	bul
opus_readme_url	https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/jpn-bul/README.md
original_repo	Tatoeba - Challenge
tags	['translation']
languages	['ja', 'bg']
src_constituents	{'jpn_Hang', 'jpn', 'jpn_Yiii', 'jpn_Kana', 'jpn_Hani', 'jpn_Bopo', 'jpn_Latn', 'jpn_Hira'}
tgt_constituents	{'bul', 'bul_Latn'}
src_multilingual	False
tgt_multilingual	False
prepro	normalization + SentencePiece (spm32k, spm32k)
url_model	https://object.pouta.csc.fi/Tatoeba-MT-models/jpn-bul/opus-2020-06-17.zip
url_test_set	https://object.pouta.csc.fi/Tatoeba-MT-models/jpn-bul/opus-2020-06-17.test.txt
src_alpha3	jpn
tgt_alpha3	bul
short_pair	ja - bg
chrF2_score	0.42200000000000004
bleu	20.2
brevity_penalty	0.9570000000000001
ref_len	2346.0
src_name	Japanese
tgt_name	Bulgarian
train_date	2020 - 06 - 17
src_alpha2	ja
tgt_alpha2	bg
prefer_old	False
long_pair	jpn - bul
helsinki_git_sha	480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535
transformers_git_sha	2207e5d8cb224e954a7cba69fa4ac2309e9ff30b
port_machine	brutasse
port_time	2020 - 08 - 21 - 14:41

📄 License

This project is licensed under the Apache - 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご