Take quality to the next level with translation model customization

modelStudio training

Create your own AI translation model with SYSTRAN Model Studio

Customization is a must-have for language professionals who need to optimize their translation workflows. Create your own customized translation model trained with your own data and increase translation productivity significantly over generic translation.

  • Available for all language professionals, no need to be an AI expert
  • Use your Translation Memories to adapt your AI translation model to your language!
  • Evaluate your model easily with automated scoring
  • Now, translate with a customized model that fits your needs!

Optimize your translation workflow with customized translation models

Use your Translation Memories to adapt your translation model to your language!

Upload your own bilingual data or Translation Memories to the platform. SYSTRAN Model Studio performs a cleaning of your data to remove duplicates and problematic segments to make it ready for a fully automated translation model training.

During the training phase, your data is processed and undergoes several transformations to ensure both quality and robustness of your translation model.

At the end, the generated translation model is used to translate a test dataset, enabling both automatic and human evaluation.

Take advantage of SYSTRAN model catalog to start your model training

Making a robust translation model from scratch requires a number of iterations and technical expertise. Luckily you won’t be starting from scratch! Benefit from SYSTRAN years of experience in translation technology, choose an existing model in our catalog as a starting point and customize it with your own data.

Our large catalog offers models across 50 language pairs and popular domains (Health, Legal, Finance, Manufacturing, Generic, etc.).

Score, evaluate and compare your model

Compare quality scores between up to three SYSTRAN models and your own customized model by adding them to the benchmark. A high score will be a good indication that your model will be good at translating your specific content.

You can either add your own test set or choose to allocate some of your training data for testing and evaluation during the data upload phase.

Data-privacy Systran

Your data remains private and secured

Data privacy and security is of utmost importance to us. That’s why we have a strict policy of not using nor exposing the data that you will upload in order to create and train your translation model.
We are relying on a secured technical infrastructure with a trusted partner, OVH, built with the highest security requirements to ensure your data is safe.


What is a translation model?

A translation model is the result of a training process with bilingual data (corpora, translation memories) that are used to make the artificial intelligence learn how to translate some content from a source language to a target language in a specific way.

Domains can be seen as large thematics, for instance News, IT, Legal, Medical… A model trained with a balance between various domains is called a “Generic” model. A model can be also tailored to perform especially well on a particular domain. This process is called specialization. To train a model that performs best on a domain, the training data should contain a majority of translation examples of that domain.

SYSTRAN Model Studio allows you to specialize a model, whether a Generic or a Domain model, and fine-tune it with your own data.

Here is an overview of our available domains for model training:

Show all available training models