I'm working on a non-English NLP project and for that, I need to re-train the UniLm on the bert-base-multilingual-cased, but they did not specify how to in their GitHub repository.
Any help is appreciated!
This is the link for the GitHub repository for Microsoft's UniLm
I hope you find this issue on github usefull, you can see if it serves your usecase. https://github.com/microsoft/unilm/issues/205