Opus-mt-en-mkh Open-source Machine Translation Model - Free English to Mon-Khmer Language Translation

Opus Mt En Mkh

Developed by Helsinki-NLP

This is a multilingual machine translation model based on the Transformer architecture, supporting translation tasks from English to various Mon-Khmer languages.

Machine Translation

Transformers

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

Downloads 58

Release Time : 3/2/2022

Model Overview

This model specializes in translation from English to Mon-Khmer languages, supporting Khasi, Khmer, Mon, Vietnamese, and other languages. It employs standardization and SentencePiece tokenization techniques for preprocessing.

Model Features

Multilingual Support

Supports translation from English to various Mon-Khmer languages, including Khasi, Khmer, Mon, and Vietnamese.

Preprocessing Optimization

Utilizes standardization and SentencePiece tokenization (spm32k) for text preprocessing.

Language Identifier

Requires adding target language identifiers (e.g., >>language_code<<) at the beginning of sentences to specify the target translation language.

Model Capabilities

English to Mon-Khmer language translation

Multilingual machine translation

Support for multiple Mon-Khmer languages

Use Cases

Language Translation

English to Vietnamese Translation

Translate English text into Vietnamese

BLEU score 33.7, chr-F score 0.513

English to Khmer Translation

Translate English text into Khmer

BLEU score 0.2, chr-F score 0.226

Multilingual Applications

Multilingual Content Generation

Generate multilingual content for Mon-Khmer language regions

🚀 eng-mkh Translation Model

This is a translation model designed to translate English into Mon - Khmer languages, offering a solution for cross - language communication within these language groups.

✨ Features

Multi - target languages: Supports translation from English to multiple Mon - Khmer languages, including Khasi (kha), Khmer (khm, khm_Latn), Mnong (mnw), Vietnamese (vie, vie_Hani).
Transformer architecture: Utilizes the Transformer model for efficient and accurate translation.
Pre - processing: Applies normalization and SentencePiece (spm32k, spm32k) for text pre - processing.
Language token requirement: Requires a sentence initial language token in the form of >>id<< (id = valid target language ID).

📦 Installation

The original weights of the model can be downloaded from opus - 2020 - 07 - 27.zip.

💻 Usage Examples

No specific code examples are provided in the original document.

📚 Documentation

Model Information

Property	Details
Model Type	Transformer
Source Language(s)	English (`eng`)
Target Language(s)	Khasi (`kha`), Khmer (`khm`, `khm_Latn`), Mnong (`mnw`), Vietnamese (`vie`, `vie_Hani`)
Pre - processing	Normalization + SentencePiece (spm32k, spm32k)
Initial Language Token	Required in the form of `>>id<<` (id = valid target language ID)
Download Link for 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

Benchmarks

testset	BLEU	chr - F
Tatoeba - test.eng - kha.eng.kha	0.1	0.015
Tatoeba - test.eng - khm.eng.khm	0.2	0.226
Tatoeba - test.eng - mnw.eng.mnw	0.7	0.003
Tatoeba - test.eng.multi	16.5	0.330
Tatoeba - test.eng - vie.eng.vie	33.7	0.513

System Info

hf_name: eng - mkh
source_languages: eng
target_languages: mkh
opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/eng-mkh/README.md
original_repo: Tatoeba - Challenge
tags: ['translation']
languages: ['en', 'vi', 'km', 'mkh']
src_constituents: {'eng'}
tgt_constituents: {'vie_Hani', 'mnw', 'vie', 'kha', 'khm_Latn', 'khm'}
src_multilingual: False
tgt_multilingual: True
prepro: Normalization + SentencePiece (spm32k, spm32k)
url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/eng-mkh/opus-2020-07-27.zip
url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/eng-mkh/opus-2020-07-27.test.txt
src_alpha3: eng
tgt_alpha3: mkh
short_pair: en - mkh
chrF2_score: 0.33
bleu: 16.5
brevity_penalty: 1.0
ref_len: 34734.0
src_name: English
tgt_name: Mon - Khmer languages
train_date: 2020 - 07 - 27
src_alpha2: en
tgt_alpha2: mkh
prefer_old: False
long_pair: eng - mkh
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