Hey Thank you for this demo! Very interesting. Currently it seems it's doesn't seem to have multilingual support (despite the OpenAI whisper-tiny model has a multilingual model). Do you happen to know if this sentis model can be swapped with a multilingual model?
Thank you for listening and your comment! I'm still an amateur at this and have no clue on how to do the mixing and just did what sounded okay to my ears, so these kind of feedback really helps! So I would presume that the high pitched guitar sounding instrument (pipa) stands out so its volume level should be lowered right?
Beautiful tracks! I love how the melodies of each track has like a little surprise moment where it didn't go the obvious way. (I hope this makes sense)
BTW the reset is not a bug but by design :) but of course if people feel like its not right I'll probably change it if I further develop this. Thanks for your feed back!
Resetting only character position and attributes was by design, but I guess it needs to be re-assessed since dying to reset became a puzzling thing... Anyhow, thank you so much for your great feedback, its super helpful!
