Thank you for the finetuned Modell

#1
by sandorkonya - opened

@johannhartmann you write in the model card that you used synthetic data for finetuning.
May i ask, which model / TTS pipeline did you use to generate the synthetic audio from the medical documents?
I tried KugelAudio but it struggles with the medical terms and the abbreviations in german. How did you overcome this?
I would like to make a finetuning of this model (and VibeVoice) myself.

Danke im Voraus!

Sign up or log in to comment