Update Swiss German Speech Data Meeting Notes authored by Michael Graber's avatar Michael Graber
## Meeting May 27, 2026
### Participants
Yixuan Xu, Daniel Perruchoud, Michael Graber
### Previous Action Items
* [ ] Roberto: Licence agreement, also partially for Regionaljournal & Espresso broadcasts (needs to be commercially usable and openly shareable)
* [ ] FHNW: share synthesized SFT QA questions via huggingface
* [ ] FHNW: evaluate synthesized SFT audio quality via ASR
### Discussion Items
* SwissDial use for TTS fine tuning for SFT data generation
* Yixuan will bring up topic in Bi-Weekly Apertus Engineering touch point
* FHNW will produce SFT dataset samples with permissively licenced voices ()
* Apertus SFT will start today
* currently only 1B audio tokens
* early capabilities are promising, but transmodal competencies are not fully visible
### New Action Points
## Meeting May 20, 2026
### Participants
......@@ -22,7 +45,7 @@ Yixuan Xu, Daniel Perruchoud, Roberto Salomone, Fabian Jeitziner, Michael Graber
### New Action Points
* [ ] FHNW: share synthesized SFT QA questions via huggingface
* [ ] FHNW: evaluate synthesized SFT audio quality via ASR
* [ ] FHNW: evaluate synthesized SFT audio quality via ASR
## Meeting May 13, 2026
......
......