Open yinjie1949 opened 1 year ago
Same issue!
same issue!
I have this issue when I run a voice model generated with serp inside bark original.
Solved. The current bark generate.py has a range of 10 voices (0-9). You need to update the range and ensure that your voices conform to the established naming convention {prefix}{lang}speaker{n}
Nice, thanks! Weird type of restriction. What good is that?
thanks very much! my problem solved.
Solved. The current bark generate.py has a range of 10 voices (0-9). You need to update the range and ensure that your voices conform to the established naming convention {prefix}{lang}speaker{n}
@alfbrown @kaka-1314 Is it possible you can elaborate a little bit
change the voice_name to such format like {lang}spreaker{n} ,such as en_speaker_0、ja_speaker_1 and so on. "en" represents English; "ja" represents Japanese.
@alfbrown @kaka-1314 Is it possible you can elaborate a little bit
but still the voice output is not the one I would like to clone
Solved!Hope it can help you
audio_array = generate_audio(text_prompt, history_prompt="/your/path/to/voice_name.npz", text_temp=0.7, waveform_temp=0.7)
I encountered this problem ValueError: history prompt not found while I execute the file clone_voice.ipynb at this part
simple generation
audio_array = generate_audio(text_prompt, history_prompt=voice_name, text_temp=0.7, waveform_temp=0.7) Please help me resolve this problem