How to apply this tool to my custom Japanese data? about multilingual_text_to_speech

[Tomiinek]

Hello, you are right. This repo contains only CSS10 and my little Common Voice subset. If you want to add your own data, you need to modify the file with meta-data and provide your audios/spectrograms. Have you seen #7 ? I think you should be ok with taking CSS10 + CV, add your custom data (including japanese CSS10) to it and retrain the model from scratch.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[zzxiang]

Thank you. I have read #7 . The result pasted on #7 (comment) is great.

Now I have roughly known how to start my own training. There may be other questions. Please let me submit them in other new issues.

Thanks again.