I tried to clone Russian speech 12 seconds and generated a recording in English. At the beginning or at the end of different recordings, there is speech in Chinese. Also, the recordings contain ringing metal. On huggingface, there is a limited quota for generation, so I could not test it further.
I tried to clone Russian speech 12 seconds and generated a recording in English. At the beginning or at the end of different recordings, there is speech in Chinese. Also, the recordings contain ringing metal. On huggingface, there is a limited quota for generation, so I could not test it further.