Looking for a voice changer, where I can train models locally or within colab

Hey there guys, I really wasn’t sure where to post this topic to be honest, but hopefully someone here can help out. I’ve seen programs like Voice.ai, and Altered Studio - but to be honest I’m not concerned with live voice changing, being limited by purchasing credits for individual voices - or paying $50 a month for the privilege. I’m wondering if there are more bare-bones/open-source versions of this concept knocking about, where I can turn a wav/mp3 of my own speech into a voice I’ve trained myself. Any searches I do lead me back to them, or dozens of text-to-speech utilities. I’m happy running Python script, or using colab notebooks - as I’ve decent experience with diffusion models and dipped my toes into deepfakery. Even if it’s more work, I like the flexibility it lends and (tbh) I’m pretty cheap.