opus-mt-en-cpp Open-source Machine Translation Model - Freely Translate English to Creole and Pidgin

Opus Mt En Cpp

Developed by Helsinki-NLP

This is a machine translation model based on the Transformer architecture, supporting translation tasks from English to various Portuguese-based Creoles and Pidgin languages.

Machine Translation

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Multilingual translation #Creole language support #SentencePiece tokenization

Downloads 29

Release Time : 3/2/2022

Model Overview

The model focuses on translation tasks from English to the Portuguese Creole language family (including variants such as Papiamento), employing standardized preprocessing and SentencePiece tokenization techniques.

Model Features

Multi-target language support

Supports translation into 7 variants of Portuguese Creole, including Standard Malay, Indonesian, Papiamento, etc.

Standardized preprocessing

Utilizes standardized text preprocessing and SentencePiece tokenization (spm32k).

Language identifier support

Enables multi-language target control by adding target language code identifiers (e.g., >>target_language_code<<).

Model Capabilities

English to Creole translation

Multi-language target output

Cross-language text conversion

Use Cases

Language services

Multilingual content localization

Translate English content into various Creole language variants to meet the needs of users in different regions.

Achieved a BLEU score of 32.7 on the Tatoeba test set.

Cross-cultural communication

Facilitate communication between English speakers and Creole speakers.

🚀 eng-cpp

This project focuses on the translation from English to Creoles and pidgins, Portuguese-based, providing a transformer model with specific pre - processing and evaluation metrics.

🚀 Quick Start

The key information about this translation model is as follows:

Source Group: English
Target Group: Creoles and pidgins, Portuguese-based
OPUS Readme: eng-cpp
Model: Transformer
Source Language(s): eng
Target Language(s): ind max_Latn min pap tmw_Latn zlm_Latn zsm_Latn
Pre - processing: normalization + SentencePiece (spm32k,spm32k)
Language Token Requirement: A sentence initial language token is required in the form of >>id<< (id = valid target language ID)
Download Original Weights: opus2m-2020-08-01.zip
Test Set Translations: opus2m-2020-08-01.test.txt
Test Set Scores: opus2m-2020-08-01.eval.txt

✨ Features

Benchmarks

testset	BLEU	chr-F
Tatoeba-test.eng-msa.eng.msa	32.6	0.573
Tatoeba-test.eng.multi	32.7	0.574
Tatoeba-test.eng-pap.eng.pap	42.5	0.633

System Info

Property	Details
hf_name	eng-cpp
source_languages	eng
target_languages	cpp
opus_readme_url	https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/eng-cpp/README.md
original_repo	Tatoeba-Challenge
tags	['translation']
languages	['en', 'id', 'cpp']
src_constituents	{'eng'}
tgt_constituents	{'zsm_Latn', 'ind', 'pap', 'min', 'tmw_Latn', 'max_Latn', 'zlm_Latn'}
src_multilingual	False
tgt_multilingual	True
prepro	normalization + SentencePiece (spm32k,spm32k)
url_model	https://object.pouta.csc.fi/Tatoeba-MT-models/eng-cpp/opus2m-2020-08-01.zip
url_test_set	https://object.pouta.csc.fi/Tatoeba-MT-models/eng-cpp/opus2m-2020-08-01.test.txt
src_alpha3	eng
tgt_alpha3	cpp
short_pair	en-cpp
chrF2_score	0.574
bleu	32.7
brevity_penalty	0.996
ref_len	34010.0
src_name	English
tgt_name	Creoles and pidgins, Portuguese-based
train_date	2020-08-01
src_alpha2	en
tgt_alpha2	cpp
prefer_old	False
long_pair	eng-cpp
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