Opus-mt-mkh-en Open-source Multilingual Translation Model - Freely Translate Mon-Khmer Languages to English

Opus Mt Mkh En

Developed by Helsinki-NLP

This is a multilingual translation model that supports translation from various Mon-Khmer languages (including Vietnamese, Khmer, etc.) to English.

Machine Translation

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Multilingual translation #Mon-Khmer language family support #Low-resource language optimization

Downloads 18

Release Time : 3/2/2022

Model Overview

This model is based on the Transformer architecture and is specifically designed for translating various Mon-Khmer languages into English, supporting multiple language variants such as Khasi, Khmer, Mon, and Vietnamese.

Model Features

Multilingual Support

Supports translation from various Mon-Khmer languages (including Vietnamese, Khmer, etc.) to English

Standardized Preprocessing

Uses standardization and SentencePiece tokenization (spm32k) for text preprocessing

Support for Multiple Language Variants

Supports translation of special language variants such as Han-Vietnamese and Latinized Khmer

Model Capabilities

Text translation

Multilingual processing

Mon-Khmer to English translation

Use Cases

Language Translation

Vietnamese to English Translation

Translate Vietnamese text into English

BLEU 34.4, chr-F 0.529

Khmer to English Translation

Translate Khmer text into English

BLEU 8.5, chr-F 0.206

Multilingual Applications

Multilingual Content Localization

Localize Mon-Khmer language content into English

🚀 mkh-eng

This project focuses on the translation from Mon - Khmer languages to English, offering a transformer - based model with specific pre - processing and evaluation metrics.

🚀 Quick Start

The mkh - eng project is designed for translating Mon - Khmer languages to English. You can download the original weights and test set translations from the provided links.

Download original weights: opus - 2020 - 07 - 27.zip
Test set translations: opus - 2020 - 07 - 27.test.txt
Test set scores: opus - 2020 - 07 - 27.eval.txt

✨ Features

Language Support: Supports multiple Mon - Khmer languages (kha, khm, khm_Latn, mnw, vie, vie_Hani) and translates them to English.
Model Type: Utilizes a transformer model for translation.
Pre - processing: Applies normalization and SentencePiece (spm32k, spm32k) pre - processing.

📚 Documentation

Model Information

Source Group: Mon - Khmer languages
Target Group: English
OPUS readme: mkh - eng
Model: transformer
Source Languages: kha, khm, khm_Latn, mnw, vie, vie_Hani
Target Language: eng
Pre - processing: normalization + SentencePiece (spm32k, spm32k)

System Info

Property	Details
hf_name	mkh - eng
source_languages	mkh
target_languages	eng
opus_readme_url	https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/mkh-eng/README.md
original_repo	Tatoeba - Challenge
tags	['translation']
languages	['vi', 'km', 'mkh', 'en']
src_constituents	{'vie_Hani', 'mnw', 'vie', 'kha', 'khm_Latn', 'khm'}
tgt_constituents	{'eng'}
src_multilingual	True
tgt_multilingual	False
prepro	normalization + SentencePiece (spm32k, spm32k)
url_model	https://object.pouta.csc.fi/Tatoeba-MT-models/mkh-eng/opus-2020-07-27.zip
url_test_set	https://object.pouta.csc.fi/Tatoeba-MT-models/mkh-eng/opus-2020-07-27.test.txt
src_alpha3	mkh
tgt_alpha3	eng
short_pair	mkh - en
chrF2_score	0.40700000000000003
bleu	24.5
brevity_penalty	1.0
ref_len	33985.0
src_name	Mon - Khmer languages
tgt_name	English
train_date	2020 - 07 - 27
src_alpha2	mkh
tgt_alpha2	en
prefer_old	False
long_pair	mkh - eng
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.

Benchmarks

Testset	BLEU	chr - F
Tatoeba - test.kha - eng.kha.eng	0.5	0.108
Tatoeba - test.khm - eng.khm.eng	8.5	0.206
Tatoeba - test.mnw - eng.mnw.eng	0.7	0.110
Tatoeba - test.multi.eng	24.5	0.407
Tatoeba - test.vie - eng.vie.eng	34.4	0.529

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