Opus-MT-AAV-EN Open-source Multilingual Translation Model - Free Deployment for Translation from Austroasiatic Languages to English

Opus Mt Aav En

Developed by Helsinki-NLP

This is a multilingual translation model based on the Transformer architecture, supporting translation tasks from Austroasiatic languages (including Vietnamese, Khmer, etc.) to English.

Machine Translation

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Austroasiatic language translation #Multi-source language support #SentencePiece tokenization

Downloads 19

Release Time : 3/2/2022

Model Overview

This model specializes in translation from Austroasiatic languages to English, supporting input from various Austroasiatic languages including Ho, Khasi, Khmer, Mon, Vietnamese, and other variants.

Model Features

Multilingual Support

Supports translation for multiple Austroasiatic languages, including Vietnamese, Khmer, and their variants.

Standardized Preprocessing

Utilizes standardized preprocessing and SentencePiece tokenization (spm32k) for text processing.

Performance Optimization

Demonstrates excellent performance on the Tatoeba test set, particularly achieving a 38.7 BLEU score for Vietnamese to English translation.

Model Capabilities

Text Translation

Multilingual Processing

Standardized Text Processing

Use Cases

Language Services

Document Translation

Translating documents from Austroasiatic languages into English

Achieves a 38.7 BLEU score for Vietnamese to English translation

Cross-language Communication

Facilitating communication between Austroasiatic language speakers and English speakers

Supports translation for multiple Austroasiatic languages

Education

Language Learning Assistance

Assisting English learners in understanding Austroasiatic language content

🚀 aav-eng

This project focuses on the translation between Austro - Asiatic languages and English, providing a transformer - based model with corresponding pre - processing and evaluation metrics.

🚀 Quick Start

This model is designed for translating from Austro - Asiatic languages to English. You can download the original weights and test set translations from the provided links for further use.

✨ Features

Multilingual Support: Supports multiple source languages from the Austro - Asiatic language group, including hoc, kha, khm, etc.
Transformer Model: Utilizes a transformer model for translation tasks.
Pre - processing: Applies normalization and SentencePiece (spm32k, spm32k) for pre - processing.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

No code examples are provided in the original document.

📚 Documentation

Model Information

Property	Details
Model Type	Transformer
Source Languages	`hoc`, `hoc_Latn`, `kha`, `khm`, `khm_Latn`, `mnw`, `vie`, `vie_Hani`
Target Language	`eng`
Pre - processing	Normalization + SentencePiece (spm32k, spm32k)
Download Original Weights	opus2m - 2020 - 07 - 31.zip
Test Set Translations	opus2m - 2020 - 07 - 31.test.txt
Test Set Scores	opus2m - 2020 - 07 - 31.eval.txt

Benchmarks

Testset	BLEU	chr - F
Tatoeba - test.hoc - eng.hoc.eng	0.3	0.095
Tatoeba - test.kha - eng.kha.eng	1.0	0.115
Tatoeba - test.khm - eng.khm.eng	8.9	0.271
Tatoeba - test.mnw - eng.mnw.eng	0.8	0.118
Tatoeba - test.multi.eng	24.8	0.391
Tatoeba - test.vie - eng.vie.eng	38.7	0.567

System Info

hf_name: aav - eng
source_languages: aav
target_languages: eng
opus_readme_url: https://github.com/Helsinki - NLP/Tatoeba - Challenge/tree/master/models/aav - eng/README.md
original_repo: Tatoeba - Challenge
tags: ['translation']
languages: ['vi', 'km', 'aav', 'en']
src_constituents: {'mnw', 'vie', 'kha', 'khm', 'vie_Hani', 'khm_Latn', 'hoc_Latn', 'hoc'}
tgt_constituents: {'eng'}
src_multilingual: True
tgt_multilingual: False
prepro: Normalization + SentencePiece (spm32k, spm32k)
url_model: https://object.pouta.csc.fi/Tatoeba - MT - models/aav - eng/opus2m - 2020 - 07 - 31.zip
url_test_set: https://object.pouta.csc.fi/Tatoeba - MT - models/aav - eng/opus2m - 2020 - 07 - 31.test.txt
src_alpha3: aav
tgt_alpha3: eng
short_pair: aav - en
chrF2_score: 0.391
bleu: 24.8
brevity_penalty: 0.968
ref_len: 36693.0
src_name: Austro - Asiatic languages
tgt_name: English
train_date: 2020 - 07 - 31
src_alpha2: aav
tgt_alpha2: en
prefer_old: False
long_pair: aav - 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.

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