Train VCTK dataset for all speakers

[SibtainRazaJamali]

Currently, vctk dataset consist of data of just 21 speakers. I want to increase number of speakers but i have not been able to see a proper guide line to prepare dataset for model training. You had just completed the formality by sharing this link
https://gist.github.com/kastnerkyle/cc0ac48d34860c5bb3f9112f4d9a0300
I have tried to execute the given file but it depends on another link
https://gist.github.com/kastnerkyle/001a58a58d090658ee5350cb6129f857
and eventually this file doesn't compile.
This code is not in usable format at all.