Model seems to have poor quality/performance on english to english voice conversion

I was trying to compare seed-vc with MeanVC hoping that MeanVC would give me higher quality results. But on english audio it seems to be far worse than Seed-vc. The  english converted audio comes out with Chinese accent even if both the source and reference audio is english.

Any chance you will provide a model that has better performance for english -> english voice conversion? 