I am planning for creating a language translator POC for excel using OpenNMT.
For excel processing I am using Pandas. For using opennmt to translate using the existing model, I am
bit confused as to how to proceed ?
One way could be this I guess, since you won’t have to deal with tokenization and detokenization, but if your PC runs on Windows, I don’t know how easy it would be to setup everything. It would be best if you could have OpenNMT in a Linux machine and just send requests there.
Another, non-pythonic, way would be to compile and use OpenNMT’s CTranslate project which doesn’t require a server.
As suggested, I have built the Linux system. How do I install the Translation Server on this .
I have the following issues:
I am unable to find the installer for Translation Server (Language translation for supported models)
The OpenNMT gives information about using lua . Foremost I do not know lua scripts.
Secondly, I am not sure how do go ahead installing using these lua scripts.
In your initial post you mentioned that you wanted to use an existing model, so I assumed you already have a working installation of OpenNMT. I suggest you read the documentation and follow the instructions about installing and using OpenNMT: http://opennmt.net/OpenNMT/
I would be using the Docker image. When I run the command
sudo docker run -it opennmt/opennmt:latest
I get the status :
Status : Downloaded newer image for opennmt/opennmt:latest