How to introduce more emotive tags?

Thanks for making this model opensource.
I have been experimenting with the model a lot and wanted to introduce more inline emotional tags like [Sneeze] [Excited] [Whisper] mid sentences to change the prosody of the speech mid-sentence as in Orpheus-TTS.

I have started by creating a synthetic dataset (tagged) -20% (12 speakers, 1000 tagged sentences, 1000 non-tagged, across 4 tags) and natural data - non-tagged- 80% (8000 sentences).
Full fine-tuning for 5 epochs. First epoch checkpoint was usable, after that became overfit.
The tags are not reliable, but previous tags are preserved and working.
The new tags are ignored, no affect at all.

Can you please share a guide on how the tags were introduced in the models or how can we go about it?

Start by reviewing the existing tag-handling path and the fine-tuning procedure described in the issue, including how prior tags are preserved. Determine what documentation or implementation guidance is needed for adding new inline prosody tags, using the reported synthetic and natural datasets as context. Done means contributors can follow a clear guide for introducing and validating new tags.