AIbase
EN
Home
Opus Mt Fr En
O

Opus Mt Fr En

Developed by Helsinki-NLP
A Transformer-based French-to-English neural machine translation model developed by the Helsinki-NLP team, trained on the OPUS multilingual dataset.
Machine Translation Supports Multiple LanguagesOpen Source License:Apache-2.0#French-English translation#Transformer alignment model#SentencePiece tokenization
Downloads 1.2M
Release Time : 3/2/2022

Model Overview

This model is specifically designed for French-to-English machine translation tasks, employing a standard Transformer architecture to support high-quality text translation.

Model Features

Transformer-based alignment model
Utilizes the Transformer architecture to support alignment between source and target languages, improving translation accuracy.
Standardized preprocessing
Input text undergoes standardization and SentencePiece tokenization to ensure consistent model processing.
Multi-test set validation
Extensively validated on multiple standard test sets, including newsdev2015 and newstest2014.

Model Capabilities

French-to-English text translation
Supports long-text translation
Batch text processing

Use Cases

Text translation
News translation
Translate French news content into English
Achieved a BLEU score of 26.2 on the news-test2008 test set
Daily conversation translation
Translate everyday French conversations into English
Achieved a BLEU score of 57.5 on the Tatoeba test set
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