Hello friends, I am using Mangiorvc to train voice with artificial intelligence. I want to train an average female voice. How long should the MP3 or WAV file be, and how should we adjust parameters like steps and epochs? I would appreciate it if those knowledgeable could provide answers.