Issue with Persian Pronunciation in Fine-Tuned Model

[mojtaba-nafez]

Self Checks

• This template is only for bug reports. For questions, please visit Discussions.
• I have thoroughly reviewed the project documentation (installation, training, inference) but couldn't find information to solve my problem. English 中文 日本語 Portuguese (Brazil)
• I have searched for existing issues, including closed ones. Search issues
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• Please do not modify this template and fill in all required fields.

Cloud or Self Hosted

Self Hosted (Source)

Environment Details

Ubuntu

Steps to Reproduce

Hi,

I fine-tuned the model using LoRA on Persian language, but encountered an issue where the model reads text with a bad accent, similar to an English speaker attempting to speak Persian. However, if the pronunciation is disregarded, the model works correctly.

Do you have any insights on why Fish Speech can't handle Persian pronunciation properly? Additionally, do you have any suggestions for resolving this issue?

Thanks!

✔️ Expected Behavior

No response

❌ Actual Behavior

No response