Adapting a dataset to nnsvs format

[linkdow]

Hi,
I was trying to adapt the SingStyle111 dataset to the toolkit in order to have other languages for training. I tried to adapt the dataset by reading the opencpop2nnsvs.py script, but got a hard time deciphering it, and ultimately failed to do so.
If someone have any insights on how to proceed or an explanation on what is the exact input format for NNSVS it would be greatly appreciated.
Thanks in advance.