Since the sample data was too little ( I extracted audio samples from the 12 episodes and two audio dramas but eventually only attained about 11 min audio available for use) , the model can’t perform perfectly in the high-pitched part of the songs, meanwhile, it works terribly in converting songs of most female singers and a few male singers with relatively high-pitched voices.
I publicly released the model weight files and index files in my GitHub and here is the link:
https://github.com/MicBao711/Hatsuragi_Maki_RVC_Model-.gitYou can directly download them and import them to ur local RVC service.