Closed alexander-rakhlin closed 1 year ago
Not in audio-webui.
I believe https://github.com/serp-ai/bark-with-voice-clone recently implemented fine-tuning of bark based on HuBERT quantizer tokens.
This has not been implemented into audio-webui, as i haven't heard such great outputs as i was hoping for yet, and i'm working on something else to add, which might sound better.
Is there general understanding how to train Bark itself?