It looks like this model or library has no control and it works arbitraty whatever it likes. Input text given and audio sample is different. Output is horrible and completely different. Same is for English or German or any other language. Is there any reason for this behavior of the model and repo?
It looks like this model or library has no control and it works arbitraty whatever it likes. Input text given and audio sample is different. Output is horrible and completely different. Same is for English or German or any other language. Is there any reason for this behavior of the model and repo?