AIbase
EN
Home
Opus Mt Ko En
O

Opus Mt Ko En

Developed by Helsinki-NLP
A Transformer-based neural machine translation model from Korean to English, supporting three Korean script variants: Hangul, Hanja, and Latin transliteration.
Machine Translation
Transformers
 Supports Multiple LanguagesOpen Source License:Apache-2.0#Mixed Hangul-Hanja Translation#SentencePiece Tokenization#Tatoeba Benchmark
Downloads 117.19k
Release Time : 3/2/2022

Model Overview

This model is specifically designed for Korean-to-English text translation tasks, employing standardized preprocessing and SentencePiece tokenization techniques, achieving a 41.3 BLEU score on the Tatoeba test set.

Model Features

Multi-Script Support
Supports input in three Korean script forms: Hangul, Hanja, and Latin transliteration
High-Quality Tokenization
Utilizes standardized preprocessing and SentencePiece tokenization (spm32k dual vocabulary)
Stable Performance
Achieves translation quality of 41.3 BLEU and 0.588 chr-F on the Tatoeba test set

Model Capabilities

Korean-to-English text translation
Multi-script Korean input processing

Use Cases

Language Services
Korean-English Document Translation
Automatically translates Korean documents into English
Achieves 41.3 BLEU score on general domain texts
Multi-Format Korean Processing
Supports translation of mixed Hangul, Hanja, and Latin transliteration inputs
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase