Incorrect accent for Indian languages

[anjalyv]

Hello,

I am very interested in using the massively multilingual Toucan TTS. I want to thank you for this incredible work. I am attempting to use Toucan for Indian languages, but the output accent is incorrect and sounds like an English accent.

• Cross checked the source language given in set_phonemizer_language() and set_accent_language() functions. It's same as my input audio
• Tested by providing reference audio.
• Fine-tuned with our data

However, the accent remains unchanged.

Am I missing something? Is there any way to improve this?