Opus-Tatoeba-en-tr Open Source Machine Translation Model - Freely Translate English to Turkish

Opus Tatoeba En Tr

Developed by Helsinki-NLP

This is a machine translation model based on the Transformer architecture for translating English to Turkish, released by Tatoeba-Challenge.

Machine Translation

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Transformer architecture #SentencePiece tokenization #News translation optimization

Downloads 689

Release Time : 3/2/2022

Model Overview

This model is specifically designed for translating English text into Turkish, utilizing standardized preprocessing and SentencePiece tokenization techniques.

Model Features

High-quality translation

Performs excellently on multiple test sets, with BLEU scores ranging from 20 to 41.

Multiple test set validation

Validated on various test sets including newsdev2016 and newstest2016-2018.

Standardized preprocessing

Employs standardized preprocessing and SentencePiece tokenization techniques.

Model Capabilities

English to Turkish text translation

Supports long text translation

Handles standardized text

Use Cases

News translation

News content translation

Translates English news into Turkish

Achieved a BLEU score of 22.8 on the newstest2017 test set

Daily communication translation

Daily conversation translation

Translates daily conversations between English and Turkish

Achieved a BLEU score of 41.5 on the Tatoeba test set

🚀 English-Turkish Translation Model

This project focuses on English - Turkish translation, offering a reliable solution for language conversion between these two languages.

🚀 Quick Start

Translation Details

Source Group: English
Target Group: Turkish
OPUS Readme: eng - tur

Model Information

Model: transformer - align
Source Language(s): eng
Target Language(s): tur
Pre - processing: normalization + SentencePiece (spm32k,spm32k)

Downloads

Original Weights: opus+bt-2021-04-10.zip
Test Set Translations: opus+bt-2021-04-10.test.txt
Test Set Scores: opus+bt-2021-04-10.eval.txt

📊 Benchmarks

Testset	BLEU	chr - F	#sent	#words	BP
newsdev2016 - entr.eng - tur	21.5	0.575	1001	16127	1.000
newstest2016 - entr.eng - tur	21.4	0.558	3000	50782	0.986
newstest2017 - entr.eng - tur	22.8	0.572	3007	51977	0.960
newstest2018 - entr.eng - tur	20.8	0.561	3000	53731	0.963
Tatoeba - test.eng - tur	41.5	0.684	10000	60469	0.932

🔧 System Info

Property	Details
hf_name	en - tr
source_languages	eng
target_languages	tur
opus_readme_url	https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/eng-tur/README.md
original_repo	Tatoeba - Challenge
tags	['translation']
languages	['en', 'tr']
src_constituents	('English', {'eng'})
tgt_constituents	('Turkish', {'tur'})
src_multilingual	False
tgt_multilingual	False
long_pair	eng - tur
prepro	normalization + SentencePiece (spm32k,spm32k)
url_model	https://object.pouta.csc.fi/Tatoeba-MT-models/eng-tur/opus+bt-2021-04-10.zip
url_test_set	https://object.pouta.csc.fi/Tatoeba-MT-models/eng-tur/opus+bt-2021-04-10.test.txt
src_alpha3	eng
tgt_alpha3	tur
chrF2_score	0.684
bleu	41.5
src_name	English
tgt_name	Turkish
train_date	2021 - 04 - 10 00:00:00
src_alpha2	en
tgt_alpha2	tr
prefer_old	False
short_pair	en - tr
helsinki_git_sha	a6bd0607aec9603811b2b635aec3f566f3add79d
transformers_git_sha	12b4d66a80419db30a15e7b9d4208ceb9887c03b
port_machine	LM0 - 400 - 22516.local
port_time	2021 - 10 - 05 - 12:13

📄 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