OpenBMB / OpenBMB/VoxCPM

How to get stable voice accent and style with HiFi cloning?

Open
#231 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
37.8k
Forks
4.3k
Avg merge
7m
Merged PRs (30d)
1

Description

I have been testing HiFi cloning. I am getting slight variations in voice depending on input text or it seems to be off on random inputs with same prompt-wav file and same prompt-wav text with cfg value of 3.

What are the options to get a stable cloned voice ? I read about LoRA fine tuning in docs for a speaker but does that mean I will have to train a new model each time I need a new voice or can multiple voices exist simultaneously in the same model ? How does that work?

Is there any other way to get a stable voice output?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

The report names no files, tests, or entry points. Begin with the HiFi cloning and LoRA fine-tuning documentation referenced in the issue, then define documentation or reproducible-test scope that explains voice stability and whether voices can coexist in one model.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
audio-video-rtc, machine-learning
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.