Is there any guide how to train a Microsoft custom MT engine and deploy it on Azure?

214 Views Asked by At

I have a big parallel corpus in TMX format which I'd like to use for training a custom Microsoft MT engine in the Microsoft Translator Hub. Then, I'd like to deploy this trained MT engine on Azure and use it in a cloud-based CAT tool. Is there any step-by-step guide how to do that?

1

There are 1 best solutions below

1
On
  • You upload your TMX to the Hub, and you train, optimize and deploy your MT engine.
  • The Hub overview page lists a category ID with each of your trained systems.
  • You copy the category ID from the Hub overview page into the "category" field that your CAT tool's MT connector exposes.

If your CAT tool does not expose a "category" field in the connector, please ask the CAT tool vendor to add it. In the Translator API call it is simply an additional parameter in the call, that allows the service to find your custom trained system when translating.