serp-ai/bark-with-voice-clone ? reverse-engineered prompt

Build me a simple local app or notebook flow that can turn text into spoken audio and also clone a voice from a short sample clip.

I want to be able to paste in some text, choose a voice preset or a custom voice sample, and generate an audio file I can play back or download as a WAV. It should also handle longer text by splitting it into chunks so the speech sounds smooth instead of cutting off. If possible, include a clean notebook or simple interface for both normal text to speech and voice cloning, with clear steps for adding a 5 to 12 second sample and reusing it for better results.

Keep it running locally, and make it easy enough that I can follow it without a bunch of setup confusion. If you need to check the latest docs for the model or related tools, go ahead and look them up online.