My finetuned model is unable to clone voices like the base f5_1200000 model.

[masooddahmedd]

Checks

• This template is only for question, not feature requests or bug reports.
• I have thoroughly reviewed the project documentation and read the related paper(s).
• I have searched for existing issues, including closed ones, no similar questions.
• I confirm that I am using English to submit this report in order to facilitate communication.

Question details

so what i want to do is that i have finetuned a model with 15 hours of a single person on Urdu Language, i have trained the model for 500k steps and it is working good. Now when i open the same model in the inference_gradio and i want to give another person's voice as reference voice. it is unable to clone that voice.

I expected that by finetuning the model we can teach it the language and later we can just clone the voices, or just train the model for another 10k steps on that new voice so it will work as the model now knows the language and only need to pick on voice characteristic, but it is unable to where as in the base model F5_1200000 it is able to clone voice.

So i have a simple question, why is the base model f5_1200000 able to clone voices but my finetuned one cant clone.

[ZhikangNiu]

Have you changed your vocab?

[masooddahmedd]

yes i have